Sexual images depicting children: the EU legal framework and online platforms' policies

Argyro Chatzinikolaou [1]


Sexual(ised) images of children may often be posted or shared on social network sites or content sharing platforms. While such material may be the result of abuse or coercion, evidence shows that it may often be linked to contemporary forms of sexual exploration or intimate communication among underage peers. The aim of this paper is to explore the boundaries of the EU legal and policy framework regulating online platforms' liability for hosting or not removing such imagery. Discussing popular online platforms' policies against their imposed responsibility to contribute to the fight against illegal child sexual abuse material (CSAM) revealed a tendency of online intermediaries to restrict more than legally required from them. However justifiable the adoption of this 'better safe than sorry' approach might be, it sparks additional controversy in relation to children's agency. Navigating between the protection and freedom of children, when the issue at stake associates with elements such as gender, morality, and culture, inherently perplexes the performed balancing and cannot guarantee easy public policy or private industry solutions. However, in the absence of clear and sufficient policy guidelines,online platforms have no other choice but to shape their policies based on popular cultural norms, their business plan, and their understanding of how sensitive content should be dealt with.

Keywords : online intermediary; liability; sexual images; children

1. Introduction

1.1 Background

Online platforms are considered great contributors to innovation and growth within the European Digital Single Market (European Commission, 2018b). Search engines, social networks, micro-blogging sites, or video- and image-sharing platforms as well as direct messaging app(lication)s mediate the access of internet users to information and digital material (European Commission, 2017b). Individuals of all ages make use of apps and websites in order to communicate thoughts and content they have created; alternatively, they navigate the online environment in quest of material and information of their interest.

In their role as facilitators of the online environment, however, online platforms may not only host legitimate exchanges, but may also serve as channels for the communication of illegal and harmful content. More precisely, internet users might express and access illegal hate speech or content that incites to terrorism, as well as access, transfer and download child sexual abuse material (CSAM)(European Commission, 2017b).

In the European Union (EU), concerns about the increasing availability of illegal content online have recently led to the adoption of non-binding guidelines and principles concerning the role of intermediaries in the fight against the online circulation of illegal content (European Commission, 2018b). Beyond the legally binding framework regarding intermediaries, which has been in place since 2000 and puts the focus on the exemption of certain categories of intermediaries from liability, and in line with the ever-growing number of users, over the last few years social media and other online platforms have been assigned particular societal responsibilities to assist this fight. Ever since 2015, the European Commission has systematically scaled up its actions towards tackling illegal content online (European Commission, 2018a) and has adopted a number of policy documents that describe the liability and responsibility of online platforms to control the content they host.

The focus of this article lays on the online availability and exchange of content that depicts children in a sexual(ly suggestive) way. Such content usually qualifies as illegal CSAM (as explicitly stated in a number of legislative and policy documents) and therefore its production, possession and dissemination are criminalised both by state- as well as EU- level legislation (see EU Directive 2011/93). In light of emerging trends of (children's) sexual exploration by means of information and communication technology - such as sexting among peers (Lanzarote Committee, 2019b; Livingstone, Mascheroni, & Staksrud, 2018; Walrave, Van Ouytsel, Ponnet, Temple, 2018) - the boundaries between the legal and illegal nature of sexual imagery depicting children are, however, increasingly blurred (Lanzarote Committee, 2019a). Sexual or sexually suggestive imagery may not always be the result of coercion or harassment; research shows that sexual images of children may be exchanged among young individuals on a consensual basis within intimate relationships or as a form of exploration of sexuality (Livingstone et al., 2018; Madigan, Ly, Rash, Van Ouytsel, & Temple, 2018; Van Ouytsel, Walrave, & Ponnet, 2018a).

1.2 Problem Statement

Whether material depicting children in a sexual way qualifies as legal or illegal cannot always be established with certainty. This results in a lack of clarity as to the applicability of existing legislation and non-binding policy documents and as to how they both shape the liability/responsibility of platforms respectively. More specifically, questions arise regarding whether an online platform should be held liable for hosting or not removing such content.

To that end, the article aims at, first to explore the possibly applicable EU legislation (and non-binding policies) on the hosting of material depicting children in a sexual(ly) suggestive way and online platforms' relevant policies. Subsequently the extent to which the platforms' approach reflects their obligations as laid down in the legislation shall be examined, and arising legal and societal controversies shall be discussed.

1.3 Methodology

In terms of methodology, the paper is based on traditional desk research. With regard to the first research question, legislation, in particular the EU Directive 2000/31 ('E-Commerce Directive'), and (non-binding) policy documents on the liability of intermediaries at a European Union level served as the basis of the conducted research. At the same time, the EU Directive 2011/93 on combating the child sexual abuse and exploitation and on child sexual abuse material contributed to delineating the liability of online platforms in relation to the material in question.

The second research question was approached through a textual analysis of the Terms and Conditions and Community Guidelines of selected online platforms on which the material in question could be made available, stored or exchanged. The selection of platforms, the policy of which was scrutinised, was based on their popularity among users as well as the reported use thereof as hosting platforms for such online exchanges. Facebook, Instagram, Snapchat, and YouTube constitute the basis of the relevant analysis.

These two parts are followed by a discussion, consisting of a critical analysis of online platforms' policies against the applicable legal framework. Subsequently, the author attempts to discuss shortcomings of the latter as well as explore nuances and issues arising from the platforms' interpretation of the law and conception of their 'duty of care' in the fight against the online circulation of child sexual abuse material. While the communication of intimate material such as sexual or sexually suggestive images (of children) raises data protection-related issues, a relevant discussion falls outside the scope of the present article. In fact, the author chooses to explore children's right to respect for private life in a broader, more principled sense, and in relation to their right to image, self-representation, and freedom of expression.

2. Liability of intermediaries with regard to sexual images of and by children - the EU legislative and policy framework

Identifying the legal landscape within which the liability of intermediaries with reference to sexual images depicting children is shaped, requires us to draw on two separate EU sets of legislative and policy instruments. On the one hand, the EU legal framework on the production, possession and dissemination of sexual images depicting children sets boundaries between legal and illegal behaviour in relation to material depicting children. On the other hand, the EU legal and policy framework on the liability of intermediaries is key to ascertaining platforms' responsibilities and liability in relation to content they host.

2.1 The EU legal framework on the production, possession and dissemination of sexual images depicting children

Imagery that depicts children in a sexual or sexually suggestive way might fall under the definition of 'child pornography', as laid down in multiple supranational as well as national legal instruments (see also EU Directive 2011/93). While this term is widely used in legally binding and non-binding policy documents, terminology considerations have been strongly voiced over the past few years. Clarification is provided for by the Interagency Working Group in Luxembourg, which adopted Terminology Guidelines for the Protection of Children from Sexual Exploitation and Sexual Abuse (widely known as 'Luxembourg Guidelines') in 2016 (Interagency Working Group on Sexual Exploitation of Children, 2016). Criticism is mainly focused on the fact that 'pornography' is increasingly normalised and may [subsequently] contribute to diminishing the gravity of, trivialising, or even legitimising what is actually sexual abuse and/or sexual exploitation of children (Frangež et al., 2016). Thus, following the Luxembourg Guidelines (pp. 38-40), the term 'child sexual abuse material' (CSAM) is used throughout this paper in replacement of the term 'child pornography'.

At the EU level, the Directive 2011/93/EU of the European Parliament and of the Council of 13 December 2011 on combating the child sexual abuse and sexual exploitation of children and child pornography (European Parliament and Council, 2011), is the relevant instrument regarding CSAM. Core aims of the legislator are the prevention and combat of sexual exploitation and sexual abuse of children, the protection of the rights of the child victims of sexual exploitation and sexual abuse and the promotion of national and international cooperation against sexual exploitation and sexual abuse of children.

According to Article 5 of the Directive, intentionally acquiring, possessing, knowingly obtaining access, distributing, disseminating, transmitting, offering, supplying, making available or producing child sexual abuse material ought to be punishable. As laid down in Article 2, 'child pornography' is defined as:

any material that visually depicts a child engaged in real or simulated sexually explicit conduct;

any depiction of the sexual organs of a child for primarily sexual purposes;

any material that visually depicts any person appearing to be a child engaged in real or simulated sexually explicit conduct or any depiction of the sexual organs of any person appearing to be a child, for primarily sexual purposes;

or any realistic images of a child engaged in sexually explicit conduct or realistic images of the sexual organs of a child, for primarily sexual purposes.

Paragraph 3 of Article 8, however, introduces a possibility for Member States to decriminalise certain behaviours that otherwise would qualify as criminal offences, insofar as certain conditions are met. More specifically, the legislator leaves it

to the discretion of Member States to decide whether Article 5(2) and (6) [making sure that the respective behaviours are punishable within the national legal framework] apply to the production, acquisition or possession of material involving children who have reached the age of sexual consent where that material is produced and possessed with the consent of those children and only for the private use of the persons involved, in so far as the acts did not involve any abuse. [2]

Notably, despite the existence of Article 8(3), only a few EU Member States have explicitly decriminalised the described behaviour; the majority of EU Member States have made no alterations to their criminal legislation, yet clarify that in practice no prosecution of children shall take place if such behaviour occurs (Chatzinikolaou & Lievens, 2020).

Finally, Article 25 was introduced in order to disrupt the availability of CSAM online. Among other, hosting service providers are considered well placed to cooperate in the implementation of Article 25 (European Commission, 2016d, p. 5), a provision underlining the need for the adoption of measures (according to Recital 47 not necessarily legislative) leading to the removal of unlawful content or the blocking of access to it. According to Article 25, [3]

Member States shall take the necessary measures to ensure the prompt removal of web pages containing or disseminating child pornography hosted in their territory and to endeavour to obtain the removal of such pages hosted outside of their territory. Member States may take measures to block access to web pages containing or disseminating child pornography towards the Internet users within their territory.

2.2 The EU Legal Framework on the liability of intermediaries

The liability of online intermediaries is regulated by the Directive 2000/31/EC of the European Parliament and the Council of the EU on certain legal aspects of information society services, and more specifically on electronic commerce in the Internal Market, widely referred to as the E-Commerce Directive (European Parliament & Council of the European Union, 2000). In addition, the European Commission demonstrates its engagement in promoting the regulation of online content and the responsibilities of platforms through two recent policy documents elaborated upon in this subsection. In line with the E-Commerce Directive and the relevant European Commission non-binding policy documents aiming to enhance platforms' responsibilities, the 2018 revision of the Audiovisual Media Services Directive (AVMSD) (European Parliament and Council of the EU) lays down measures to be taken by video-sharing platform providers with regard to content they may host, without prejudice to Articles 12 to 15 of the E-Commerce Directive (Article 28b of the AVMSD).

2.2.1 E-Commerce Directive

According to the preamble of the E-Commerce Directive, its goal was to create a basic legal framework which would remove obstacles to cross-border online services in the EU and provide legal certainty to business and citizens in cross-border transactions. To that end, certain categories of intermediaries are exempted from liability resulting from the illegal nature of the information provided by the recipient of the service under certain conditions that are laid down in Articles 12 - 15 of the E-Commerce Directive.

The E-Commerce Directive applies to 'information society services', meaning 'any service normally provided for remuneration, at a distance, by electronic means and at the individual request of a recipient of services' (pursuant to the definition of Article 1(2) of Directive 98/34/EC as amended by Directive 98/48/EC). Pursuant to Article 2(b) of the E-Commerce Directive, a 'service provider' is 'any natural or legal person providing an information society service'. According to recent case law of the Court of Justice of the EU (CJEU), 'it is common ground' that social media platforms - such as Facebook - 'provide the services of a host provider for the purposes of Article 14 of Directive 2000/31' (Judgment of 3 October 2019, Eva Glawischnig-Piesczek v Facebook Ireland Limited , C-18/18, EU:C:2019:821, paragraph 22).

Articles 12-14 do not lay down legal grounds for the establishment of liability; on the contrary, they describe the circumstances under which exemption from liability applies. The nature of the provisions is horizontal in the sense that (a) they cover different types of illegal content and activities (such as content that infringes defamation laws, provisions on the protection of minors, privacy laws or commercial practices) and (b) they cover different kinds of liability (both criminal or civil and direct or indirect) (Le Borgne-Bachschmidt et al., 2008, p. 220). The liability exemptions offered through articles 12, 13 and 14 apply to intermediaries providing 'mere conduit' (Article 12; an information society service is provided that consists of the transmission in a communication network of information provided by a recipient of the service, or the provision of access to a communication network), 'caching' (Article 13; an information society service is provided that consists of the transmission in a communication network of information provided by a recipient of the service), and 'hosting' services (Article 14; an information society service is provided that consists of the storage of information provided by a recipient of the service). As already stated above, the CJEU has by now established that social network platforms qualify as hosting intermediaries and therefore can benefit from the liability exemption of the provision (see C-18/18; Judgment of 16 February 2012, SABAM v Netlog , C-360/10, EU:C:2012:85, paragraph 27).

Moreover, pursuant to Article 15,

Member States shall not impose a general obligation on providers, when providing the services covered by Articles 12, 13 and 14, to monitor the information which they transmit or store, nor a general obligation actively to seek facts or circumstances indicating illegal activity.

This basic principle allows for Member States to delineate and specify duties of care that are expected from intermediaries in their role to protect and prevent illegal activities, as described in recital 48 (Koelman & Julia-Barcelo, 2000). Despite the explicit reference to 'duties of care', applicable solely for the prevention and detection of certain types of illegal activities foreseen by national law, neither the Directive nor the CJEU elucidate the meaning of this concept (Kuczerawy, 2019, p. 2). While Article 15 and the extent of the monitoring obligations that may be imposed on platforms by Member States remains a debatable point in need for further clarification (Kuczerawy, 2019), the CJEU recently provided for an interpretation of the provision. In the Eva Glawischnig-Piesczek v Facebook Ireland Limited judgment (C-18/18, paragraph 53), the Court ruled that

Article 15(1), must be interpreted as meaning that it does not preclude a court of a Member State from: ordering a host provider to remove information which it stores, the content of which is identical [or equivalent, when additional circumstances are met] to the content of information which was previously declared to be unlawful, or to block access to that information, irrespective of who requested the storage of that information; and from ordering a host provider to remove information covered by the injunction or to block access to that information worldwide within the framework of the relevant international law,

somewhat stretching the conception of monitoring that may be imposed on platforms. Anyhow, tracking and removing content which is identical to content that has been found to be unlawful is distinct from monitoring for unlawful content per se.

Given the focus of the article on online platforms on which sexual imagery depicting children might be stored and made available to users, the analysis is centred on Articles 14 and 15 and the liability regime for hosting intermediaries.

According to Article 14 of the E-Commerce Directive,

information service providers whose (a) service consists of the storage of information (hosting service) and who (b) have no actual knowledge of illegal activities or information (or, in case of damages, of facts or circumstances from which the illegal activity or information is apparent) or who, upon such knowledge act expeditiously to remove or to disable access to the information (c) are not liable for information stored at the request of a recipient of that service.

Through Article 14, the E-Commerce Directive introduces a notice-and-take-down/action mechanism, the efficacy, proportionality and adequacy of which has been extensively debated upon in literature (Kuczerawy, 2018a). As described in the provision, once a hosting service has knowledge of infringing activities, it must act 'expeditiously' to remove or to disable access to the infringing information.

2.2.2 Audiovisual Media Services Directive (AVMSD)

While it does not regulate the liability of intermediaries per se, the AVMSD contributes to the realisation of the 'enhanced liability regime' (European Commission, 2017b) for online platforms (Montagnani, 2019). In particular, after the 2018 review, Article 28b, without prejudice to Articles 12 to 15 of the E-Commerce Directive, calls Member States

to ensure that video-sharing platform providers under their jurisdiction take appropriate measures to protect [among other] the general public from programmes, user-generated videos and audiovisual commercial communications containing content the dissemination of which constitutes an activity which is a criminal offence under Union law, namely [..] offences concerning child pornography as set out in Article 5(4) of Directive 2011/93/EU of the European Parliament and of the Council.

2.2.3 Non-binding EU policy documents on the responsibility of online platforms in tackling illegal content online

In parallel to the E-Commerce Directive, the European Commission has drafted a number of complementary non-binding documents that provide guidance to online service providers and Member States on the role of the former as hosts of content generated by third parties. In fact, the European Commission has recently been promoting the enhancement of their 'responsibilities' in dealing with illegal third party content (European Commission, 2015).

Firstly, in 2016, the European Commission drafted a Communication on Online Platforms and the Digital Single Market (European Commission, 2016a). Secondly, the Communication on Tackling Illegal Content Online, issued in September 2017, includes a 'set of guidelines and principles for online platforms to step up the fight against illegal content online in cooperation with national authorities, Member States and other relevant stakeholders' (European Commission, 2017b). The Communication 'aims to facilitate and intensify the implementation of good practices for preventing, detecting, removing and disabling access to illegal content so as to ensure the effective removal of illegal content, increased transparency, and the protection of fundamental rights online'. Furthermore, it is meant to provide clarifications to platforms on their liability when they take proactive steps to detect, remove or disable access to illegal content (the so-called 'Good Samaritan' actions).

Thirdly, building on this Communication, the Commission issued a Recommendation on measures to effectively tackle illegal content online (European Commission, 2018b), in which once again CSAM is explicitly singled out. Among others, it is proposed that 'hosting service providers be encouraged to take, where appropriate, proportionate and specific proactive measures in respect of illegal content'. Such proactive measures may, for instance, involve the use of automated means for the detection as well as the prevention of the dissemination of such material (European Commission, 2018b, p. 6). Hosting intermediaries are further encouraged to collaborate closely with Member States, trusted flaggers (meaning an individual or entity which is considered by a hosting service provider to have particular expertise and responsibilities for the purposes of tackling illegal content online), as well as other hosting service providers (European Commission, 2018b, p. 13) in order to effectively remove or disable the access to illegal content.

3. Online Platforms' Policies on Sexual(ly Suggestive) Material Depicting Children

3.1 Introduction

Material that depicts children in a sexual or sexually suggestive way may be found on or exchanged through different types of online platforms, be it search engines and portals, or participative networking platforms - such as social network sites, instant messaging or file-sharing platforms (Sartor, 2017, p. 7). The posting or distribution of such imagery - including videos - may be initiated by different parties and for a number of different purposes. To begin with, (adult) perpetrators create, make available and distribute child sexual abuse material; the digital depiction of child sexual abuse is widely communicated within the so called 'dark web', through websites that are not accessible via standard browsers (Kobie, 2019). However, the focus of this article will limit itself to online platforms that operate on the regular web and are considered to play a key role in facilitating the digital economy and society and consequently the EU Digital Single Market (European Commission, 2016a). The sending of self-made sexually explicit [text messages], images, and videos through the computer or the mobile phone, as sexting is defined (Van Ouytsel, Walrave, Ponnet, & Temple, 2018) , may occur among individuals under the age of 18 years old on a consensual basis as a way to maintain or establish romantic relationships (Burén & Lunde, 2018). In other words, children engage in consensual sexting by using popular social media platforms and direct messaging apps (Van Ouytsel, Van Gool, Walrave, Ponnet, & Peeters, 2017). It goes without saying, though, that such platforms may be also engaged by third persons for the unauthorised forwarding or posting of material that was initially created consensually by the depicted child. Finally, naked images of children that portray their sexual organs are also posted on social media profiles by parents who wish to casually and innocently share moments of their offspring's childhood (Steinberg, 2017).

Against this background, the Terms and Conditions of a number of online platforms were looked into. More precisely, their policy on allowing users to 'post' online sexual images (either still or in motion) of children constitutes the core of the analysis. The selection of online platforms was based on their nature as possible hosts of sexual imagery depicting children, in combination with their popularity among the internet users - be it children themselves or adults - who may engage in the distribution of the content in question, as portrayed in either scholarship or the media(Anderson & Jang, 2018; Crofts, Lee, McGovern, & Miliovojevic, 2018; Elgersma, 2019; Moreau, 2019; Ofcom, 2019). On the basis of these two selection criteria, the following participative networking platforms (photo- and video-sharing sites and social network sites) were chosen: Facebook (Elgersma, 2019; Ofcom, 2019, p. 8), Instagram (Ofcom, 2019, p. 8), Snapchat (Charteris & Gregory, 2018; Piwek & Joinson, 2016), and YouTube (reportedly becoming the viewing platform of choice among children)(Ofcom, 2019, p. 5).

3.2 Online Platforms' Terms and Conditions

The relevant policies were searched for under the terms 'Community Standards' (Facebook), and 'Community Guidelines' (Instagram, Snap Inc, YouTube). First, the author looked for the platforms' rules on the posting of child sexual abuse material, sexualised images of children (that do not represent the sexual abuse of children, see Luxembourg Guidelines, p. 42), and nude imagery depicting children. Second, the sanction applied in case of a breach of the platform's policy as well as the possibility for users to report restricted content which has been posted on the platform were both looked into.


The policy of Facebook on sexual images depicting children can be found under the section 'Community Standards' on the platform's website (Facebook, n.d.-a). More precisely, under the subsection on 'Safety' one can find both the policy rationale and examples of restricted content in relation to 'Child Nudity and Sexual Exploitation of Children'. As described, content that is sexually exploiting or endangering children is not allowed on Facebook. Drawing from the indicative list provided, Facebook users are not to post content that depicts participation in the sexual exploitation of children, content (including photos, videos, real-world art, digital content and text) that depicts - among other - any sexual activity among minors, or minors with sexual elements (including focus on genitals) nor content that shows minors in a sexualised context.

Beyond material that represents children in a sexual or sexually abusive/exploitative manner, Facebook also restricts content that depicts child nudity, meaning genitals (even when covered or covered by transparent clothing), straight and/or fully bare back at close range, unshielded female nipples for toddlers, children without clothes, from neck to knee, who are older than toddlers as well as digitally generated images of naked minors, unless the image is intended for health or educational purposes. As explicitly explained by the platform, even though it is understandable that 'sometimes people share nude images of their own children with good intentions, [Facebook] generally removes these images because of the potential for abuse by others and to help avoid the possibility of other people reusing or misappropriating the images'(Facebook, n.d.-a).

According to the 'Safety Center' section of the platform, in case a user would like to report the existence of content on the platform which violates the Community Standards, Facebook includes a link on nearly every piece of content for reporting abuse, bullying, harassment and other issues. When the platform becomes aware of apparent child exploitation, it reports it to the United States (US) National Center for Missing and Exploited Children (NCMEC). [4]

Moreover, it is worth mentioning that according to the platform's community standards, nudity and the representation of sexual activity involving adults is also restricted. Within the section titled 'Objectionable Content' and under the subsection on 'Adult Nudity and Sexual Activity', Facebook explains that the display of nudity or sexual activity is restricted all in all, as 'some people in our [the Facebook] community may be sensitive to this type of content', and in order to prevent the sharing of non-consensual or underage content(Stewart, 2017, p. 171). In this case, however, the threshold between acceptable or non-acceptable adult nudity may vary pursuant to cultural and social norms(Kuklis, 2019, p. 9). As stated explicitly by the platform, their 'nudity policies have become more nuanced over the time', allowing for exceptions to the prohibition of nude or sexual content sharing involving adults.


The policy of Instagram (which has been owned by Facebook since April 2012) on sexual images depicting children can be found on the 'Privacy and Safety Center' page of the platform's website, a subsection of Instagram's Help Center, (Instagram, n.d.). Instagram offers guidance to users with regard to 'pornography' involving children under the section 'Child Exploitation'. In a 'Question and Answer' format, the question 'What should I do if I see images that sexually exploit children (example: pornography) on Instagram?' is followed by the explicit warning that 'posting, sharing or downloading images that sexually exploit children for any reason can be criminal' as well as that the platform 'reports all apparent child pornography to the National Center for Missing and Exploited Children'. Users are subsequently requested not to share or comment on such material, to report it to the platform through the offered 'built-in reporting options' (which are included in the list appearing on the Privacy and Safety page) and to notify the NCMEC.

The basic principles by which the Instagram 'community' members must abide are laid down in the platform's 'Community Guidelines'. Instagram has 'zero tolerance when it comes to sharing sexual content involving minors or threatening to post intimate images of others'. Furthermore, as explicitly described, 'Instagram is a reflection of our diverse community of cultures, ages and beliefs' and therefore users may 'post photos and videos that are appropriate for a diverse audience'. According to the Guidelines, while it is acknowledged that 'there are times when people might want to share nude images that are artistic or creative in nature, [..] for a variety of reasons', nudity is not allowed on Instagram. The platform goes on to provide for examples of restricted material, including 'photos, videos, and some digitally-created content that show sexual intercourse, genitals, and close-ups of fully-nude buttocks or some photos of female nipples'. Yet, 'photos of post-mastectomy scarring and women actively breastfeeding are allowed'.

While the abovementioned policy of nudity restriction seems to make no distinction based on whether the depicted person is an adult or a child, Instagram further acknowledges the phenomenon of 'sharenting', a term describing how parents share details about their children's lives online (Steinberg, 2017). Users are warned 'that there are times that for safety reasons pictures showing nude or partially nude children may be removed, even when shared with good intentions'. As explained, their policy seeks to prevent 'the use of such content by others in unanticipated ways'. Parents are then directed to the Instagram's 'tips for parents page', where the reasons why an image of one's child posted by the parent may be removed from the platform are listed.

In case a user breaches the platform's Guidelines, according to the Terms of Use, Instagram may 'removeany content or information shared on the Service if [they] believe that it violates these Terms of Use, policies (including our Instagram Community Guidelines) or [they] are required to do so by law'. Alternatively, in the case of 'clear, serious or repeated' violation thereof, Instagram 'can refuse to provide or stop providing the service to the user'.


The policy of Snap Inc. (the company that has developed Snapchat) on imagery that depicts children in a sexual way can be found under the section 'Community Guidelines' and 'Snap Inc. Terms of Service' (Snap, n.d.). As regards the Terms of Service, a distinction is made based on whether the users live in or outside the United States. In the latter case, and under the subtitle 'Safety', the company explicitly states that '[they] try hard to keep their services safe' and subsequently explain that by using the services, individuals 'agree [among other] that they will not post content that contains pornography'. Moreover, Snap Inc. puts forward a number of Community Guidelines 'to support [their] mission by encouraging the broadest range of self-expression while making sure Snapchatters can use our services safely every day'. To that end, a separate reference is made to 'sexually explicit content'. At a first level, Snap Inc. 'prohibits accounts that promote or distribute pornographic content' and goes on to clarify that 'breastfeeding and other depictions of nudity in non-sexual contexts are not considered pornographic and are permitted on Snapchat' (Snap, n.d.). At a second level, according to the Guidelines, users should 'never post, save, or send nude or sexual content involving anyone under the age of 18 — even of oneself [and also] never ask a minor to send explicit imagery or chats'.

The company generally states that child sexual exploitation is reported to the authorities, without however providing additional, more detailed information on the process and their collaboration with the competent authorities. On the same page, users are informed of the possibility offered to them at all times to file a report with the company's safety team either by using the in-app reporting feature or by completing a form, both of which are directly provided for through hyperlinks. As a matter of fact, users are encouraged to visit the website's 'Privacy Center' where the company's safety policy is further elaborated.

Snap Inc. reviews the filed reports and the safety team in charge determines whether a violation of either the Terms of Service or the Guidelines has occurred. In such a case, the offending content may be removed, the account may be terminated and/or the law enforcement authorities may be informed.


The policy of YouTube (now operating as a subsidiary of Google) on sexual images depicting children can be found on the 'Policies and Safety' page. One can then be easily directed to the platform's 'Community Guidelines', and specifically the guidelines on 'nudity and sexual content', the 'Safety Tools and Resources' and information on how to report content on YouTube along with information on how YouTube enforces their Community Guidelines. As explained both in written form and through a video, YouTube 'is not meant for pornography or sexually explicit content'. YouTube provides for a long indicative list of content types/examples that are not allowed on the platform, or that shall be otherwise age-restricted. The 'depiction of genitals, breasts, or buttocks (clothed or unclothed) for the purpose of sexual gratification [as well as] pornography depicting sexual acts, genitals, or fetishes for the purpose of sexual gratification' are not allowed.

Furthermore, within their policy on 'Child Safety on YouTube', it is made clear that 'sexually explicit content featuring minors and content that sexually exploits minors is not allowed on YouTube'. Examples of such material would be a video featuring minors engaged in provocative, sexual, or sexually suggestive activities, challenges and dares, such as kissing or groping ora video that advertises sexual content featuring minors. Users are warned that content containing child sexual abuse imagery are reported to the US NCMEC (who work with global law enforcement agencies). In fact, users are repeatedly encouraged to 'flag' content that (possibly) violates the Community Guidelines, through the report button offered on every YouTube video. Alternatively, the platform offers additional options for reporting, including the 'Reporting Tool', which is appropriate for a more detailed and extensive report for review.

YouTube staff review reports on a 24/7 basis; once it is determined that the content in question violates the Guidelines, the content is removed and the individual is notified accordingly. The platform 'has zero tolerance for predatory behaviour on YouTube' and informs users that assistance is provided to law enforcement agencies in investigations, insofar as it is believed that a child is in danger.

4. Discussion

4.1 Information Society Services Providers, Intermediaries, Online platforms - definitional confusion

A first finding relates to the definitional confusion caused by the use of different terms across policy documents and legislative instruments with reference to notions that seem to (at least partly) overlap. The E-Commerce Directive makes reference to information society services providers (for instance, Article 2) or intermediary service providers (for instance, recital 45), while scholars simply use the term '(online/internet) intermediaries' (Frosio, 2017; Kuczerawy, 2018b). More recently, however, a shift has been made towards the use of the term online platforms (Bevolder, 2019; European Commission, 2015; Sartor, 2017), a term quite loosely defined (Taddeo & Floridi, 2017, p. 327). Despite the absence of consensus on a possible legal definition of online platforms (European Commission, 2016b), in its Communication on online platforms (2016), the European Commission indicatively lists online advertising platforms, marketplaces, search engines, social media and creative content outlets, application distribution platforms, communications services, payment systems, and platforms for the collaborative economy as online platforms.

The E-Commerce Directive is considered the key legislative instrument for the regulation of their operation and liability regime (European Commission, 2016a, 2017b, 2018b), yet the lack of a coherent and consistent definition may cause further confusion in relation to the scope and applicability of legislative and policy instruments.

4.2 Is the liability exemption of Article 14 of the E-Commerce Directive applicable to online platforms on which sexual images depicting children may be stored?

While the application of the E-Commerce Directive to all platforms discussed above is widely acknowledged (European Commission, 2017b, 2018b; Le Borgne-Bachschmidt et al., 2008; Sartor, 2017), whether they benefit from the liability exemption offered in Article 14 is less straightforward. Looking back at the wording of the provision reveals key elements that determine when the platform may benefit from the liability exemption; the platform ought to qualify as a 'hosting information society service provider', the activity or information stored on the platform ought to be illegal, and the platform ought to have no actual knowledge of illegal activity or information that they store or upon obtaining such knowledge acts expeditiously to remove or to disable access to the information. Yet, the issue of when these conditions are met has been an object of dispute in academia as well as in case law ever since the introduction of the E-Commerce Directive, and especially after the emergence of new types of online platforms over the past decade (Kuczerawy, 2018; Koelman & Julia-Barcelo, 2000; Le Borgne-Bachschmidt et al., 2008; Sartor, 2017; van Hoboken, 2012). Ever since the E-Commerce Directive entered into force, the European Commission has recurrently assessed the adequacy and implementation of the Directive and the liability regime put forth, including through the launch of a public consultation on the Regulatory Environment for Platforms, Online Intermediaries andthe Collaborative Economy (European Commission, 2017a; Frosio, 2017). Whilst the E-Commerce Directive still remains the key regulatory instrument for the liability of intermediaries, the debate on its adequacy shall be reopened at a legislative level through a new Digital Services Act (which 'will upgrade liability and safety rules for digital platforms, services and products') announced by the European Commission's President-elect in her Agenda for the new term of the European Commission (von der Leyen, 2019, p. 13).

4.2.1 Nature of the intermediary: passive or active?

Applying the liability exception of Article 14 of the E-Commerce Directive presupposes a clear understanding of its scope, with the nature of the intermediary being a crucial element to consider. While the concept of a hosting service provider is at first sight delineated in the E-Commerce Directive itself through Article 14 and Recital 42, what types of intermediaries qualify as hosting intermediaries is strongly debated upon in literature (Kuczerawy, 2018b; Van Eecke & Truyens, 2014).

For now, light has been shed on the scope of Article 14 through the case law of the CJEU (Bevolder, 2019, p. 23); the Court has ruled that social media platforms, such as Facebook, ( SABAM v Netlog , C-360/10; Eva Glawischnig-Piesczek v Facebook Ireland Limited , C-18/18) as well as search engines ('in the case where that service provider has not played an active role of such a kind as to give it knowledge of, or control over, the data stored'; Judgment of 23 March 2010; Google France SARL and Google Inc. v Louis Vuitton Malletier SA and Others, C-236/08 to C-238/08, EU:C:2010:159) fall within the meaning of 'hosting service providers' of Article 14.

As set out in the E-Commerce Directive (Recital 42), hosting services are only services whose activities are 'of a mere technical, automatic and passive nature, which implies that the information society service provider has neither knowledge of nor control over the information which is transmitted or stored'. On the basis of the expansive application of Recital 42 by the CJEU, a distinction between passive and active hosting has been established. After the CJEU rulings in Google France and Google v Luis Vuitton Malletier a.o ., Joint Cases C-236/08 to C-238/08 (paragraphs 113-114) and L'Oréal v eBay , Case C-324/09 (paragraphs 112-116), the exemption of Article 14 applies to hosting intermediaries as long as their role is restricted to being sufficiently passive. The restrictive nature of Article's 14 interpretation has attracted abundant criticism (Kuczerawy, 2018c; Sartor, 2017; Patrick Van Eecke, 2011). In fact, according to Van Eecke (2011), as opposed to Recital 43 (which describes not being involved in any way with the transmitted information as a condition for applying the liability exemption to mere-caching and conduit services), benefiting from the liability exemption of Article 14 does not require a passive role on behalf of the platform. The so called 'storage but no knowledge test' draws a line for benefiting from the liability exemption on the basis of not having actual knowledge over the data stored (Kuczerawy & Ombelet, 2015).

In practice however, liability immunity is conditioned upon passivity on behalf of the platform, the practical meaning of which remains blurred. On the one hand, taking an increasingly active role regarding the way in which platforms offer their services to users may fall within a platform's freedom to conduct business, which has been repeatedly acknowledged in relation to hosting service providers by the CJEU (see, for instance, SABAM v Netlog, C-360/10, paragraph 42; Sartor, 2017, p. 22). Contributing to framing the way in which user-created content is created or accessed (through, for instance, indexing, prioritising or linking) is usually of instrumental value for the main function of the platform. Social networks, for instance, decide on which content receives prominence in a user's feed or suggest links to users on the basis of similar interests. In a similar way, search engines and portals index the content that has been uploaded by users in a way that would make their navigation easier and more attractive, still without actually creating the content itself. In other words, a certain degree of initiative or editorial discretion is necessary for an intermediary to carry out core functions to its users (Yoo, 2012).

On the other hand it remains unclear whether the adoption of proactive measures render intermediaries active, and therefore obstruct them from benefiting from the liability exemption (Kuczerawy, 2018c). While the distinction between active and passive intermediaries still holds, the recent policy documents published by the European Commission, namely the Communication on Tackling Illegal Content Online and the Recommendation on measures to Effectively Tackle Illegal Content Online, set forth the enhancement of online platforms' responsibilities and their role in fighting illegal content online. As KUCZERAWY aptly notes, the relevant policy discourse has steadily shifted from liability (relating to a negligence-based approach) to responsibility (emphasising the need for proactive measures) of intermediaries (2019). The European Commission is requiring a more active role from online platforms, which should still be in conformity with the prohibition of introducing a general monitoring obligation - as set out in Article 15 of the E-Commerce Directive. Yet at the same time, it explicitly states that 'the mere fact that an intermediary hosting service provider takes certain measures relating to the provision of its services in a general manner does not necessarily mean that it plays an active role in respect of the individual content items it stores'. As expressed by legal scholars, whether the CJEU 'would agree that employing proactive measures to detect and remove illegal content as an enforcement of the terms of service is similarly passive as merely setting the terms of service' is highly questionable (Kuczerawy, 2018c).

A clarification on how wide the liability exemption is, which would explicitly expand its scope as to cover 'active' intermediaries insofar as they are clearly not the creators of the content, is timely and much awaited (Kuczerawy, 2018c). In the author's opinion, the sensitivity of material that illustrates children in a sexual manner, coupled with the great uncertainty as to whether allowing it on the platform triggers liability, are sufficient grounds to justify the 'better safe than sorry' approach adopted by both popular search engines and participative networking platforms.

4.2.2 Nature of the material: illegal or legal?

The nature and business model of the online platforms listed above may vary, however they appear to have both a common purpose - the sharing and dissemination of content to the public (French Government, 2019) - and a quite uniform approach towards hosting sexual imagery depicting children.

For online intermediaries to be held liable for user-created content they may host, the digital material in question needs to be classified as illegal pursuant to the applicable national legal framework. Whereas the determination of the illegality of a certain type of content often requires a case-by-case assessment against the legal framework in place, there seems to be absolute consensus across EU Member States that child sexual abuse (CSAM) and child sexual exploitation material (CSEM) constitutes illegal content, and therefore behaviour relating to the material ought to be punishable (see Directive 2011/93 and the criminalisation of the intentional production, possession, acquisition and dissemination of child sexual abuse material, Article 5). Without exception, the policy section of all four platforms makes explicit reference to the prohibition of CSAM - often referred to by the term 'child pornography' - and CSEM. Against this background, the uniform approach detected across online platform, according to which child sexual abuse material is prohibited, is aligned with the legislation in place.

At a second glance, though, one realises that determining what constitutes (illegal) child sexual abuse material is in fact not as straightforward. In fact, EUROPOL has acknowledged how challenging it is to reliably discriminate between consensual and non-consensual types of behaviours in cases of 'youth-produced sexual content' (2017). The increasing trend of creating and distributing such material, meaning 'nude or semi-nude images or videos produced by a young person of themselves engaging in erotic or sexual activity and intentionally shared by electronic means' (Internet Watch Foundation, 2015, p. 3),as well as Article 8(3) of the Directive 2011/93, blur the line between legal and illegal.

The adopted definition of child sexual abuse material in the Directive 2011/93 is the starting point for defining the type of content platforms are required by law not to host. According to Article 2, the definition of child sexual abuse material includes any material that visually depicts a child engaged in real or simulated sexually explicit conduct or any depiction of the sexual organs of a child for primarily sexual purposes .

A few observations should be made on the legal qualification of sexual imagery involving children. Firstly, despite the broadness of the definition [5] - at least at EU level - the depiction of a child's sexual organs constitutes illegal child sexual abuse material insofar as it occurs for 'primarily sexual purposes'. It emerges from the definition included in the Directive 2011/93, that the context in which the images are taken as well as the intention of the actor(s) play a crucial role in the legal determination of the material. Subsequently, two compelling questions arise: firstly, does 'youth-generated sexual content' constitute illegal material?; and secondly, can naked pictures of children uploaded on social media profiles by parents or by themselves (Ouvrein & Verswijvel, 2019) qualify as child sexual abuse material?

i) 'Youth-generated sexual content'

To begin with, the possibility offered to Member States to decriminalise the production, possession or acquisition of material which could fall within the scope of the definition of CSAM on the condition that the children involved have reached the age of sexual consent, the material is produced and possessed with the consent of those children and only for the private use of the persons involved, in so far as the acts did not involve any abuse (Article 8 par. 3 of Directive 2011/93), challenges the qualification of sexual images of children as illegal under certain circumstances. Thus far, the approach to sexual or sexually suggestive material depicting children has been mainly protectionist towards underage individuals (Bulger et al., 2017). However, it is by now a fact reflected in social science research, that children create, possess and exchange sexual(ised) images or videos of themselves on/through social media platforms and direct messaging apps as part of the exploration of their sexual identity or as an expression of sexuality (De Ridder, 2017; Livingstone & Mason, 2015; Walrave, Van Ouytsel, Ponnet, & Temple, 2018; Van Ouytsel, Walrave, & Ponnet, 2019). This reality was recognised by the EU legislator and is reflected in Article 8(3) of the Directive 2011/93. Recital 20 clarifies that consensual sexual activities in which children are involved may be regarded as the normal discovery of sexuality in the course of human development, involving new forms of establishing and maintaining relations among children and adolescents, including through information and communication technologies, and are not meant to be regulated by the Directive. While sexting is by now widely perceived as legitimate behaviour performed by underage individuals, whether it also constitutes legal behaviour remains a matter of debate across Europe; research shows that a limited number of countries have made use of the legislative exception, while many more tend to interpret their unaltered criminal legislation towards the same direction (Chatzinikolaou & Lievens, 2020).

It has been argued that 'self-produced digital content captures a sexual 'private moment' and potentially turns it into a public one' (Quayle & Cariola, 2019). Regardless of one's adherence (or not) to such a statement, the reference to 'private' and 'public' moments brings to the fore an important concept: the public or private nature of the act and/or the space which accommodates it. Decriminalising the production, possession or acquisition of content in cases where children engage in consensual sexting is more profoundly linked to the exchange of imagery through direct messaging applications (which often offer the option of encryption) and on a one-to-one basis. When discussing about online communication channels such as Facebook profiles, or Instagram and Snapchat accounts, the notion of online private or public space ought to be revisited. The classification of a space, communication channel or act may influence the level of privacy expected (Timan et al., 2017), however it is broadly accepted in scholarship that it ought not be assumed that everything digital is automatically public (Nissenbaum, 2011; Solove, 2007). Moreover, accepting that existing privacy norms for off-line social interactions help us assess what digital content should be deemed private (Nissenbaum, 2011)rationally leads to the assertion that the digital capture of sexual or sexually suggestive behaviour is also private (Hasinoff, A. & Shepherd, T., 2014). Acknowledging that a consensual online sexual act carried out by children in a (more) private virtual space does not (necessarily) lead to harm and may constitute a contemporary way of exploring one's sexual identity has brought along a policy shift (Chatzinikolaou & Lievens, 2019). While social network profiles are accessible to a larger audience compared to a peer-to-peer - more confidential - communication, the account holder still remains in control of what information is available to who in the first place (Hasinoff, A. & Shepherd, T., 2014) by choosing to set up a private over a public account. The account holder, nonetheless, may not fully control the further circulation of the material they produced. A parallel, though, could be drawn between private messaging spaces (which traditionally accommodate sexting) and social networking profiles accessible to multiple digital 'friends'. Against this background, the question rises whether the condition of 'private use' (laid down in the legislative exception of Article 8(3) of the Directive) is still met when a child chooses to post a sexual or sexually suggestive image for their online friends to see. It further leads to the question of whether the risks associated with the wider accessibility of intimate imagery allow for such an interpretation of the legislative exception. If interpreted broadly, the material would fall outside the scope of criminalisation (always depending on the transposition of both Articles 2 and 8(3)of the Directive in the national legislation), and would not qualify as illegal CSAM; subsequently the online platform would not be legally required to prohibit it.

ii) Nudity

The online platforms above, however, do not only restrict content portraying children in a sexual or sexually suggestive way, but also explicitly oppose nude imagery of children in general. Whilst some platforms are more illustrative of the type of content considered acceptable (or not) through examples, others suffice to merely refer to the prohibition of nudity, sexual or pornographic imagery. Facebook, for instance, provides a detailed description of what the definition of nudity includes, while Snap does not allow nude imagery of children taken by oneself.

Capturing nude parts of a body may not always carry a sexual element, which is necessary for the material's classification as illegal CSAM. It might often be the case that parents post and share online imagery from their offspring's (especially young children or babies) private life (Ouvrein & Verswijvel, 2019; Steinberg, 2017; Verswijvel et al., 2019). Alternatively, without having a primarily sexual purpose in mind, children themselves often post images that reveal their (intimate) body parts. Images taken during summer vacations on the beach or by the pool or artistic photography that captures the naked body (Ferenc, 2018), however,may not pass the threshold of acceptable content set by the social network sites. Quite interestingly, in the absence of a sexual intent on behalf of the creator and distributor of the material, and considering the rationale behind the criminalisation of the production, possession or dissemination of child sexual abuse material (which is combating the abuse of children) as well as the wording of the exception, such pictures of children taken and made public online by parents or themselves do not qualify as illegal content. [6]

The realisation of such an inconsistency between legislative imperative and private industry policy accents the often opaque line between legitimate and legal acts and unavoidably extends the discussion to socio-cultural perceptions of morality. Inevitably, the acceptability or acceptance of nudity in spaces perceived as public also becomes of interest. Creating and distributing naked images of oneself online is often perceived as a 'technological, sexual, and moral crisis' (Hasinoff, 2013). In fact, the level of displayed nudity that constitutes risk for children has not yet been sufficiently explored in literature (Tariq et al., 2019). By posting sexy or otherwise body-revealing selfies, young individuals are seen as engaging in intimate edgework and thus negotiating the societal boundaries of morality/immorality (Hart, 2017).

In addition to the restriction of sexual and nude imagery involving children, platforms' policies also often prohibit the sexual as well as nude representation of adults. Popular participative networking platforms tend to forbid sexual or nude pictures of adults, irrespective of whether the image involves themselves. In most cases, platforms appear to regularly revise their policy conforming to social norms and the ethical expectations of their users, providing them with more updated, detailed, and nuanced guidelines (Kuklis, 2019). As described in Facebook's Community Guidelines on nude pictures and adult sexual activity, 'viewing naked or sexual activity is not allowed because some members of our community may be sensitive to such content'.

Contemporary western cultures tend to contextualise the exposure of bodily nakedness and legitimise it by excluding the sexual element from certain sites or contexts (such as communal showers or bathing children) (Cover, 2003). However, the definition of 'indecent' material, which may vary across national legal frameworks, is decisive for the legal classification of the material. According to the United Kingdom Sentencing Guidelines Council, 'there may be cases where an image is not posed or 'erotic' but could still be deemed indecent, for example, a naked picture of a child not engaged in sexual activity but with a focus on the child's genitals' (2013, p. 80). In general, exceptions to the nudity restriction approach usually relate to a medical, educational, humorous, artistic or satirical context.Indicatively,Facebook allows for exceptions when images are posted as a protest, to inform the public for a certain matter, or for educational or medical purposes; Instagram allows nudity in photos of paintings or sculptures; and YouTube allows nudity 'when the primary purpose is educational, documentary, scientific, or artistic, and it isn't gratuitous'. In particular, as highlighted by YouTube, the context in which nudity is portrayed plays a crucial role in its qualification as acceptable material. It may be the case, alternatively, that the content is categorised as 'age-restricted' rather than being restricted from the platform all in all (see YouTube - material including nudity or sexual depiction may also qualify as age-restricted material).

At the same time, the restriction seems to be upheld, or at least debated, in cases where the lawfulness of the activity is (generally) not disputed. The most prominent examples are the widespread prohibition of the representation of female nipples, and the (non-)qualification of images illustrating females breastfeeding as indecent and restricted material (Locatelli, 2017).

In the former case, the restriction of the visual representation of females' nipples on social media has sparked an intense public debate on the lopsided treatment of females and males (Acosta, 2016). The nudity ban, which exclusively affects women, has triggered public outrage, demonstrations and campaigns, such as the 'Free the Nipple' movement; and despite being brought to the fore a few years ago, the debate still appears timely (Paul, 2019).

In the latter case, conflicting perceptions were voiced on the application of the nudity restriction rule by Facebook on images showing females breastfeeding (Khan, 2008). The discourse relates to a broader debate on whether it is acceptable for females to breastfeed in public (Amidi, 2017; Doward, 2018), and led to a policy shift on behalf of some social network platforms, including Facebook and Instagram (Gander, 2014). Albeit users report that in practice the posting of such imagery may still trigger their ban from the Service (Sampathkumar, 2017), major social networking sites explicitly exempt breastfeeding images from their nudity ban (see Community Guidelines of Facebook and Instagram).

Subsequently, it is made apparent, that a double standard exists when assessing acceptable nudity (in the digital sphere). Based on one's gender, specific body-parts (and subsequently activities and poses) may take on a sexual connotation; more specifically, while a picture of male nipples arouses no moral controversy, capturing female breasts almost intrinsically carries a sexual element or is otherwise classified as indecent. Moreover, mediated female body parts are often commodified and receive negative attention in the context of gendered discourses over sexuality and morality (Ringrose et al., 2013). Thus, dispatching the sexual element from the nude representation of a body highly depends on social expectations of modesty attached to each gender.

Nudity in the online sphere may reflect and reproduce moral norms and discourses of offline public nudity. In fact, the restrictions on nudism or public nudity, which appear to be quite mainstream across European states, link such behaviour to criminalised indecent exposure, public indecency, sexual exhibition, disturbances to public order, and threats to public peace (de Vries, 2019). Similarly, even if imagery depicting a naked body of a child may not be deemed child sexual abuse material, it might still qualify as illegal material depending on the societal and legislative tolerance to public nudity.

As the line between legality and illegality in the examples demonstrated above is considerably thin, online platforms find themselves in the difficult position of assessing the (legal) nature of such pictures and determining whether allowing them or not blocking access to them would lead to being held liable. In light of this reality, and taking into account the legitimate risk of misuse of content by third parties (see, for instance, Van Ouytsel et al., 2017), the absolute restriction of sexual, sexually suggestive or nude imagery involving children is considered by the author a reasonable tactic on behalf of the platforms. The rationale behind the restriction policies seems to be common across platforms. Despite pictures being shared or posted with the best of intentions, either by guardians or legal representatives of children or by the children themselves, nude or semi-nude pictures of children are to be removed owing to the potential of abuse and in order to avoid being used or owned by others (see Facebook Guidelines on child nudity and sexual exploitation of children) (O' Neill, 2018).

Beyond this reasoning, child imagery depicting nudity or carrying a sexual element is in principle removed in order to prevent intimate material being shared without authorisation by the depicted person as well as being perpetually circulated online (Gillespie, 2011). While the participation of children in social networks is associated with opportunities and positive experiences, digitally mediated online exchanges between children may also carry a number of (safety) risks (Livingstone, Davidson, & Bryce, 2017; Livingstone, Ólafsson, & Staksrud, 2013; Van Ouytsel et al., 2018). Even when content has been initially communicated on a consensual basis and no abuse has taken place, the unauthorised distribution of the material which has been acquired by that person - either consensually or illegally - is possible. Such exchanges may lead to secondary victimisation of the child, the adverse consequences of which have been widely acknowledged in literature over the past two decades (Europol, n.d.; Gillespie, 2011, p. 205; Taylor & Quayle, 2003, p. 194; Van Ouytsel et al., 2017, p. 449).

In light of the ambiguity around the measures they ought to take for tackling illegal content online (mainly caused by the lack of detailed and more pragmatic guidance offered to platforms by the European Commission) and the subsequent responsibility they carry pursuant to the most recent European Commission policy documents (which will be further discussed later on), restricting any digital expression of sexuality or nudity in respect of children is a nuanced solution. In addition, the absence of a single international definition delineating what constitutes CSAM (Gillespie, 2011)perplexes the task of legally classifying material posted online and accessed by users across different countries. Platforms tend to avoid making the legal assessment of controversial content which may cause them liability. Meanwhile, as the users of such social media platforms come from very diverse cultural backgrounds, finding a single set of rules which would accommodate all social norms and nuances seems quite unrealistic (Kuklis, 2019, p. 9). A case-by-case evaluation of whether a sexual image of a child sent through the platform to a peer constitutes legal material would be welcome, yet appears difficult. Platforms have opted out from evaluating the content's (il)legality and adopt a safer - allegedly for both themselves and children - approach by not allowing any sexual/nude images depicting children on the platform. However reasonable such a strategy may be, though, it triggers further controversy. As will be analysed more thoroughly below, applying a blanket ban on any material depicting children nude or in a sexually suggestive manner without assessing its legality could easily be read as collateral censorship (Sartor, 2017).

4.2.3 Actual vs constructive knowledge

Knowledge is usually acquired by platforms through the 'notice-and-action' report mechanisms in place as well as through the use of automated (algorithmic) content filtering mechanisms (see, for instance, Alghowinem, 2019). A hosting intermediary may benefit from the liability exemption of Article 14 as long as it did not have 'actual knowledge of illegal activity or information' and, with reference to claims for damages, has not been 'aware of facts or circumstances from which the illegal activity or information is apparent' (meaning 'constructive knowledge'). Upon obtaining such knowledge or awareness, the service provider is further required to act expeditiously to remove, or disable access to, the information.

Yet, once again, a crucial condition for determining whether an intermediary shall be held liable or not for hosting illegal content has sparked criticism. Article 14 establishes a dual standard for the required level of knowledge when it comes to criminal and civil liability respectively; actual knowledge of illegal activity or information is required for the determination of criminal liability, while 'constructive knowledge' is required for the acknowledgement of civil liability. The boundary between the two types of acquired knowledge may also be blurred(Kuczerawy, 2018b, p. 62). In fact, national courts still face difficulties interpreting and applying the threshold of actual knowledge in practice (see, for instance, the Request for a preliminary ruling from the Bundesgerichtshof (Germany) lodged on 6 November 2018 — LF v Google LLC, YouTube Inc., YouTube LLC, Google Germany GmbH,C-682/18).

According to the ruling of the CJEU in L'Oreal v eBay , 'apparent illegality occurs when any diligent economic operator should have identified the illegality in question' (paragraph 120). Along the same lines, during the review process of the E-Commerce Directive, it was proposed by certain stakeholders, that intermediaries should be only required to remove 'only manifestly illegal content' (European Commission, 2012).

Regardless of how reasonable such a proposal might be in general terms (Kuczerawy & Ausloos, 2015, p. 242), its application to the context of sexual images of children posted or exchanged online may not be as unencumbered. Even though sexual images of children are usually considered CSAM and therefore constitute illegal material, as elaborated above, exceptions to this rule might apply. Again, in light of the uncertainty around the legal qualification of sexual images depicting children, whether the platform has actual or even constructive knowledge shall not always be sufficiently clear nor easy to determine.

4.3 Online platforms' fight against the circulation of Child Sexual Abuse Material: international cooperation, enforcement mechanisms and intrinsic limitations

In his very recent report for the Council of Europe, CARR accents the need for multi-disciplinary collective action at national and international level to prevent and combat child sexual exploitation and abuse material (2019). Cooperation between public and private actors is similarly promoted by the European Commission (Jaeger, 2019). As a range of stakeholders are required to take action, the online platforms make their engagement in fighting against the (sexual) exploitation and abuse of children explicit and repeatedly express 'zero tolerance' towards it or towards those who post, save or distribute the material (see the Community Guidelines of Instagram, Snapchat and YouTube). At the same time, leading technology companies, among which Facebook and Google, support the 'We Protect Global Alliance to End Child Sexual Exploitation Online' (European Commission, 2016c), the alliance that resulted from the merge of 'We Protect' (a UK-driven global multi-stakeholder initiative for combating online child abuse and exploitation, established in 2014) and 'The Global Alliance Against Child Sexual Abuse' (an alliance launched by the European Commission and the US in 2012) (Baines, 2019; WeProtect Global Alliance, 2016). However, as social networks transcend geographical national borders, no matter how necessary international cooperation is, the intrinsic limitations related to both jurisdictional issues (French Government, 2019, p. 13) and the co-existence of multiple legal frameworks that may not be fully coherent shall not be overlooked.

It is common practice for platforms to enable users to report material and (other) users that violate the community guidelines or terms of use of the service (Council of Europe, n.d.). As a matter of fact, all platforms offer the possibility to report content that violates the terms of use of the service. Facebook, Instagram, Snap Inc., and YouTube inform their users that upon review of the report, which is carried out by the competent team appointed by the platform, the finding of a violation results in the 'taking-down' of the material. In addition, it leads to the seizure of one's profile and the communication of the violation to law enforcement insofar the behaviour also violates the applicable legislation. Without exception, all above-mentioned platforms refer to their close collaboration with Law Enforcement Agencies (LEA) or (supra)national centres for reporting child exploitation or abuse. Individuals may be further updated on the prevalence of such violations and the amount of material that was taken down by the platform through the annual reports on the enforcement of the community standards, issued by some platforms. Facebook, for instance, regularly issues reports on the enforcement of the platform's guidelines; according to the latest report, the platform has actioned 5.1 million pieces of content relating to child nudity and sexual exploitation over the first quarter of 2019 - a number that demonstrates decrease in comparison to 2018 (Facebook, n.d.-b). The efficiency of their 'notice-and-take down' policies is, however, often questioned due to the vast amount of illegal content that remains accessible on the platform (Alexander, 2019).

Reliably identifying and classifying CSAM is significantly challenging for law enforcement agencies (Europol, 2017; Kloess et al., 2018), let alone for private enforcement mechanisms. Not only is it necessary, yet sometimes hard, to determine whether the image depicts a child (Mayer et al., 2014), but, most importantly, the decision of whether the image is of an indecent nature (Kloess et al., 2018) may spark additional controversy. According to Gillespie, identifying clearly indecent images may be considerably easy, yet defining less explicit images (for instance, identifying the context of an image depicting a partially clothed adolescent) is far more difficult (2011). Pornographic or nude image recognition mechanisms employed by social media platforms rely mainly on trained algorithms (Brennan & Phippen, 2020; Ion & Minea, 2019; Shayan et al., 2015) and human moderators (Van Royen et al., 2016). As part of their risk management strategies and their contribution to the fight against the circulation of illegal content - including CSAM - private intermediaries engage algorithmic enforcement, thus far employed by law enforcement authorities (Elkin-Koren & Perel, 2018). Training, however, an artificial intelligence to tell apart nudity may not be easy (Stephen, 2019). Moreover, in borderline situations where a sexual focus cannot be directly identified and contextual details play a pivotal role in determining whether the material is indecent or not, subjective judgment is necessary (Kloess et al., 2018, p. 177). In such occasions, relying on deep learning for classifying imagery may fall short of the perplex assessment to be carried out (Brennan & Phippen, 2020). Existing content filtering solutions for the detection of explicit nudity and pornography, based for instance on the percentage of skin regions in an image, may not suffice for more context-specific cases (Ion & Minea, 2019). It thus comes as no surprise that algorithmic content filtering solutions may erroneously label content as illegal and subsequently remove it from the platform. In fact, such a margin of error is indirectly acknowledged even by Facebook. As explained in the November 2019 Community Standards Enforcement Report, while appeals on content that has been actioned for sexual exploitation material are not allowed, they were exceptionally introduced 'for child nudity violations - such as innocent images of children in the bath'.

To conclude, while intermediaries seem well positioned to engage in the fight against the circulation of illegal CSAM by performing content regulation, the nature of the content in question makes their task of classifying material depicting children in a sexual way more challenging and the classification methods deployed vulnerable to criticism.

4.4 Balancing rights: Excessive collateral censorship vs legitimate restriction of freedom of expression and the right to a private life vs freedom to conduct business

It results from the above, that online platforms tend to restrict more than is strictly required from them by the legislator. Even though nude or sexual images of children may not necessarily constitute illegal CSAM, the knowing hosting of which would trigger the platform's liability, popular social network and content sharing platforms tend to ban such content entirely and regardless of the specific circumstances under which it has been created and shared. In this section, the focus is put on the posting of imagery by the children themselves, rather than by third persons, such as their parents. [7]

Whether the enforced restrictions are proportionate merits further exploration, which could in fact benefit from the European Data Protection Supervisor Guidelines on assessing the proportionality of measures that limit the fundamental rights to privacy and to the protection of personal data (2019). Based on the analysis offered in the Guidelines, measures undertaken by hosting intermediaries ought to be necessary and appropriate (proportionate stricto sensu) in light of the legitimate aim pursued in order to be held proportionate in a broad sense (Albinati, 2019, p. 8).

The online platforms' nudity and sexual content restriction approach may be justified partly by the binding and non-binding policy regime on intermediary liability, which sparks considerable legal uncertainty, and partly by their freedom to conduct business and to shape their platform's profile within the digital sphere. In either case, though, there does exist a high risk of collateral censorship and (disproportionate) interference with the platform user's freedom of expression and right to private life . Performing a balancing exercise might prove quite challenging, yet reflecting on the competing rights already brings to the fore possible shortcomings of the legal framework and/or the private industry's initiatives. Contemporary forms of intimate peer interaction (sexting) (Smahel et al., 2020, p. 82) or self-expression through image creation (be it sexually suggestive, artistic or merely informative of one's activities and daily life) call for a reconsideration of existing policies.

Without doubt, online platforms play a crucial societal role in enabling communication and exchanges between peers in the digital environment and safeguarding internet users' ability to freely express and explore their identity (Heins, 2013). It has by now been both established in literature (Livingstone et al., 2018; Walrave, Van Ouytsel, Ponnet, & Temple, 2018) and acknowledged at a policy level (see Directive 2011/93 Article 8(3); Chatzinikolaou & Lievens, 2020), that the engagement of children in the practice of sexting amounts to a legitimate form of sexual exploration. As children create and explore new forms of expression and intimate communication with peers through social media platforms, it is not only the loss of freedom of expression they fear of, but also the limitation of their intertwined right to privacy (van der Zwaan et al., 2014; Van Royen et al., 2016, p. 39). According to literature and international case law [8], intimate decisions, primarily of a sexual nature (Koops et al., 2016), as well as activities that occur in both public and private spaces and encompass sensitive issues such as [..] sexual preferences (Finn et al., 2013) fall within the rather elusive concept of privacy. Sharing intimate imagery or texts of sexual nature remains vital for an individual's self-development and self-determination (Koops et al., 2016), and so platforms' users maintain the right to protect personal information from being monitored. Besides the international protection of the individuals' right to freedom of expression and right to private life (European Convention on Human Rights, Articles 10 and 8, and Charter of Fundamental Rights of the European Union, Articles 11 and 7), children merit special protection on the basis of the United Nations Convention on the Rights of the Child. In particular, pursuant to Articles 6(2), 13 and 16, children hold the right to determine their self-presentation by exercising control over their image, as their rights to development, freedom of expression, and privacy are guaranteed. Besides discourses around agency, the choice of making sexualised or nude images of oneself accessible to peers is often attributed to sexualised mass media representations (Hasinoff, 2013),and double standards around female (and in particular underage girls') sexuality have been detected by a number of scholars (Cooper et al., 2016; Ringrose et al., 2013; Tolman, 2012). Within feminist scholarship, debates still exist over the digital representation of female bodies and sexuality, and over whether it results in the sexual objectification of the female body or flags the sexual liberation and agency of girls (Ringrose et al., 2013). Nonetheless, the creation and communication of sexualised imagery by adolescents is undoubtedly associated with (risky) opportunities (Livingstone et al., 2017).

Restricting nudity and sexual representations involving children could be partly underpinned by the enhanced responsibilities enforced upon online platforms by the European Commission as part of its strategy to actively involve intermediaries in the fight against illegal content online. Despite the publication of the Recommendation on measures to effectively tackle illegal content online (2018) and the Communication on Tackling Illegal Content Online (2017), what is expected from the online platforms remains quite vague. The way in which they interpret the assumed duties of care in protecting children is to some extent left to their discretion or, in some countries, is elucidated by the national courts' (detailed) evaluation of the preventive measures or overall duty of care that can be imposed on platforms (see, for instance, Germany) (Ullrich, 2017). How they may accomplish their role as contributors to the fight against CSAM highly depends on the business model, the technological and economic capacity of the intermediary, and the perception of the ethical responsibilities to be met towards society as a whole and the platforms' (potential) users in particular (Sartor, 2017, p. 31; Taddeo & Floridi, 2017).

As stated in literature, the legal quality and the social value of legal regulations imposing sanctions on certain types of communications (in our case, sexual images depicting children), shall be evaluated on the basis of both the extent to which harmful, illegal content is removed or blocked and on the extent to which 'it does not deter lawful communication' (Sartor, 2017, p. 12). Objections to the posting of a nude or sexual(ly suggestive) image of a child where no abuse has taken place may be justified by a legitimate fear of misuse by third parties (Quayle & Cariola, 2019; Wolak et al., 2018). Nonetheless, the line between legitimate restriction of freedom of expression and unwanted censorship is thin (Kuczerawy, 2018; Sartor, 2017).

Allowing social network platforms and other digital platforms to restrict such imagery could on the one hand be a welcomed interpretation of platforms' imposed responsibility to tackle illegal content online (Myers West, 2018). Yet a ban as broad as those described above may arguably be seen as the 'privatisation of censorship' (Etzioni, 2019; Heins, 2013, p. 325) and thus be unwanted. The 'self-regulated' removal and blocking models adopted by private companies are reasonably criticised as resulting in censorship 'without the usual constitutional safeguards that apply to public authorities' (McIntyre, 2011; Nair, 2019, p. 107).

In parallel, social network and content sharing platforms are meant to allow for users to share content from their everyday lives, be it images of a child itself or one's child on the beach, images from a child's christening or the depiction of art representing a nude child which one wishes to post online.

To the extent that platforms choose to restrict content which constitutes a legitimate act of personal (sexual) expression, they may be partially failing to strike a fair balance between their societal role as online communication facilitators and protectors of children.

5. Conclusions

Sexual images of children can be found posted online on content sharing or social media platforms. While it could often be the case that the material is the result of sexual abuse or coercion of the depicted child, evidence shows that imagery that depicts children in a sexualised way may as well constitute a contemporary form of exploration of sexuality and sexual communication performed by the children themselves. Alternatively, the posting of imagery featuring nude body parts of children may be void of any (underlying) sexual element; it might be parents willing to share private family moments including a somewhat more revealing representation of their underage offspring or children who have chosen to share content including (semi) nudity.

The popular online platforms looked into, namely Facebook, Instagram, Snapchat and YouTube, seem to adopt a quite uniform approach towards imagery depicting children in a sexual way; not only child sexual abuse material, but also nude representations of children are restricted. As noticed, no platform in the analysed sample appears to make any distinction between content posted by a third person or the depicted child itself nor assesses the material based on the existence (or not) of a sexual purpose. Similarly, specific contextual circumstances seem to play a limited role in the classification process.

The adopted policies partly result from the legal obligations set out by the binding and non-binding EU regime on intermediaries' liability and on the combat against CSAM. However, the platforms' policies reflect a tendency to restrict more than is legally required from them. Central to this ascertainment is, in fact, the legal definition of CSAM and the following qualification of content as illegal. Though the determination of what constitutes illegal CSAM might be challenging owing to the cross-border operation of such intermediaries and the diverse definitions adopted by national legislators, at the EU level and pursuant to the Directive 2011/93 there needs to exist a primarily sexual purpose behind the depiction of a child's sexual organs or the capture of explicit sexual conduct.

This active policy choice could be justified by a number of reasons. Firstly, the wording of Article 14 of the E-Commerce Directive, the key provision of the EU intermediary liability regime which affords a conditioned liability exemption to hosting intermediaries, creates uncertainty as to its scope and applicability. Similarly, parallel to the EU tendency to enhance online platform's responsibility in tackling illegal content online, both the E-Commerce Directive and the non-binding policy instruments issued by the European Commission allow online platforms a margin to adopt measures that meet the imposed responsibility. The uncertainty as regards the triggering of liability, if the platform abstains from removing or blocking the access to sexual material depicting children, rationalises the 'better safe than sorry' risk management strategy taken up by the intermediaries. In other words, the lingering unclarity and legal uncertainty justifiably causes over-enforcement and highlights the need for and significance of additional policy guidance.

This absolute restriction of online nudity and sexual representation of children is intrinsically linked to social, cultural, and moral anxieties (Lee et al., 2013)and the consequent challenge in distinguishing indecent (or harmful) content from what ought to be labelled as socially acceptable.

While a sexualised media culture that influences the social norms of underage peer (intimate) communication and self-representation is widely acknowledged, what is considered 'age-appropriate' behaviour is still debated across different countries and different cultures. The classification of intimate material depicting children may be largely subjective and immensely challenging in light of diverse societal and moral narratives and the importance of contextual circumstances. Due to such innate difficulties, the efficiency of automated filtering enforcement mechanisms deployed by private industry cannot but come into question. And no matter how understandable the 'better safe than sorry' approach is, it inevitably gives rise to further controversy and the burning question: Is the interference with children's right to privacy, self-representation and freedom of expression proportionate/proportionately legal? Filtering and monitoring content (by automatic means or human resources) might be desirable as pursuing a legitimate aim, yet safeguarding children's autonomy rights in the digital environment is pivotal (Council of Europe, 2018; Van Royen et al., 2015). Navigating between protection and freedom of children when the issue at stake associates with elements such as gender, morality, and culture, inherently perplexes the performed balancing and cannot guarantee easy public policy or private industry solutions. As aptly put by Ringrose, Harvey, Gill, and Livingstone, 'What would it mean for us to live in a world where teen girls could unproblematically take, post or send an image of their breasts to whomever they wished?' (2013).

With the minimum age of criminal responsibility ranging from 10 to 16(European Union Agency for Fundamental Rights, 2018a)and the age of sexual consent usually set between 14 and 16 across the EU(European Union Agency for Fundamental Rights, 2018b), one may wonder why underage individuals are not allowed to capture their naked body (be it in a sexualised way or not) and make it accessible to their online peers. Thus far, notwithstanding the legal nature of the activity portrayed, the intentions behind it, the context of the imagery, and children's awareness of the risks associated with sharing intimate content online, it seems that the limited ability to control what happens next overrides children's agency rights. Whether our society can afford and is willing to accept nude or sexualised imagery depicting children online, in an environment baring risks, pertains to our conceptual understanding of (online) 'public space' and our tolerance towards intimate forms of self-expression coming from underage individuals.

In any event, so long as the EU regime on intermediary liability is not revised to accommodate the way in which online intermediaries operate today and no sufficient guidelines on the measures to be taken by online platforms in their fight against illegal sexual imagery of children exist, online platforms have no other choice but to shape their policies based on popular cultural norms, their business plan, and their understanding of how sensitive content should be dealt with.


Acosta, D. (2016, January 26). The Fight to Free the Nipple. The New York Times . Retrieved from

Albinati, F. (2019). EDPS Guidelines on assessing the proportionality of measures that limit the fundamental rights to privacy and to the protection of personal data.

Alexander, J. (2019, February 19). YouTube still can't stop child predators in its comments. Retrieved from The Verge website:

Alghowinem, S. (2019). A Safer YouTube Kids: An Extra Layer of Content Filtering Using Automated Multimodal Analysis. In K. Arai, S. Kapoor, & R. Bhatia (Eds.), Intelligent Systems and Applications (pp. 294-308).

Amidi, F. (2017, July 30). President's daughter in breastfeeding row. BBC News . Retrieved from

Anderson, M., & Jang, J. (2018, May 31). Teens, Social Media & Technology 2018 | Pew Research Center.

Baines, V. (2019). Online child sexual exploitation: Towards an optimal international response. Journal of Cyber Policy , 4(2), 197-215.

Bevolder, B. (2019). The Platform Economy—Unravelling the Legal Sta t us of Online Intermediaries (1sted.). Retrieved from

Brennan, M., & Phippen, A. (2020). Child Protection and Safeguarding Technologies: Appropriate or Excessive 'Solutions' to Social Problems? (1sted.). Routledge, New York. Retrieved from

Broadcasting Authority of Ireland. (2019). Submission to the Government's Public Consultation on the Regulation of Harmful Content and the Implementation of the Revised Audiovisual Media Services Directive.

Bulger, M., Burton, P., O'Neill, B., & Staksrud, E. (2017). Where policy and practice collide: Comparing United States, South African and European Union approaches to protecting children online. New Media & Society , 19(5), 750-764.

Burén, J., & Lunde, C. (2018). Sexting among adolescents: A nuanced and gendered online challenge for young people. Computers in Human Behavior, 85, 210-217.

Carr, J. (2019). Mechanisms for collective action to prevent and combat online child sexual exploitation and abuse—A comparative review . Council of Europe. Retrieved from

Charteris, J., & Gregory, S. (2018). Snapchat and digitally mediated sexualised communication: Ruptures in the school home nexus. Gender and Education .

Chatzinikolaou, A., & Lievens, E. (2019). Towards a Legal Qualification of Online Sexual Acts in which Children are Involved: Constructing a Typology. European Journal of Law and Technology, 10(1).

Chatzinikolaou, A., & Lievens, E. (2020). A legal perspective on trust, control and privacy in the context of sexting among children in Europe. Journal of Children and Media, 14(1), 38-55.

Cooper, K., Quayle, E., Jonsson, L., & Svedin, C. (2016). Adolescents and self-taken sexual images: A review of the literature. Computers in Human Behavior, 55, 706-716.

Council of Europe. (2018). Recommendation CM/Rec(2018)7 of the Committee of Ministers to member States on Guidelines to respect, protect and fulfil the rights of the child in the digital environment.

Council of Europe. (n.d.). Reporting on Social Media Platforms. Retrieved 24 October 2019, from No Hate Speech Youth Campaign website:

Crofts, T., Lee, M., McGovern, A., & Miliovojevic, S. (2018). Sexting pleasures: Young people, fun, flirtation, and child pornography. In Liam. Grealy, Catherine Driscoll, Anna Hickey-Moody, Youth, Technology, Governance, Experience: Adults Understanding Young People (pp. 103-122). Retrieved from

Cover, R. (2003). The Naked Subject: Nudity, Context and Sexualization in Contemporary Culture. Body & Society , 9(3), 53-72.

De Ridder, S. (2017). Social Media and Young People's Sexualities: Values, Norms, and Battlegrounds. Social Media + Society , 3(4), 2056305117738992.

de Vries, B. (2019). The Right to be Publicly Naked: A Defence of Nudism. Res Publica , 25(3), 407-424.

Doward, J. (2018, September 16). I'm glad I started a debate, says athlete who breastfed on ultra-marathon. The Guardian . Retrieved from

Elkin-Koren, N., & Perel (Filmar), M. (2018). Algorithmic Governance by Online Intermediaries (SSRN Scholarly Paper ID 3213355). Social Science Research Network.

Elgersma, C. (2019, June 6). 18 Social Media Apps and Sites Kids Are Using Right Now. Retrieved from

Etzioni, A. (2019). Should We Privatize Censorship? (SSRN Scholarly Paper No. ID 3464349). Retrieved from Social Science Research Network website:

European Commission. (2012). Commission Communication to the Europ ean Parliament, the Council, the Economic and Social Committee and the Committee of Regions—A coherent framework for building trust in the Digital Single Market for e-commerce and online services (No. COM(2011) 942 final). Brussels

European Commission. (2015). Online Platforms . Retrieved from Digital Single Market—European Commission website

European Commission. (2016a). Communication on Online Platforms and the Digital Single Market Opportunities and Challenges for Europe . Retrieved from

European Commission. (2016b). Full report on the results of the public consultation on the Regulatory environment for Platforms, Online Intermediaries and the Collaborative Economy . Retrieved from

European Commission. (2016c). We Protect Global Alliance to End Child Sexual Exploitation Online . Retrieved from Migration and Home Affairs — European Commission website

European Commission. (2016d). Report from the Commission to the European Parliament and the Council assessing the implementation of the measures referred to in Article 25 of Directive 2011/93/EU of 13 December 2011 on combating the sexual abuse and sexual exploitation of children and child pornography . Retrieved from

European Commission. (2017a). Archive — E-commerce directive — What happened before and since its adoption . Retrieved from Digital Single Market—European Commission website

European Commission. (2017b). Communication on Tackling Illegal Content Online — Towards an enhanced responsibility of online platforms . Retrieved from Digital Single Market — European Commission website

European Commission. (2018a). A Europe that protects: Countering illegal content online . Retrieved from

European Commission. (2018b). Recommendation on measures to effectively tackle illegal content online [Text]. Retrieved 14 March 2019, from Digital Single Market—European Commission website:

European Parliament and Council. (2011). Directive 2011/92/EU of the European Parliament and of the Council of 13 December 2011 on combating the sexual abuse and sexual exploitation of children and child pornography, and replacing Council Framework Decision 2004/68/JHA . Retrieved from

European Parliament and Council of the EU. (2010). Directive 2010/13/EU on the coordination of certain provisions laid down by law, regulation or administrative action in Member States concerning the provision of audiovisual media services (Audiovisual Media Services Directive). Retrieved from

European Parliament, & Council of the European Union. (2000 ). Directive 2000/31/EC of the European Parliament and of the Council of 8 June 2000 on certain legal aspects of information society services, in particular electronic commerce, in the Internal Market ('Directive on electronic commerce'). Retrieved from

European Union Agency for Fundamental Rights. (2018a). Children's rights and justice - Minimum age requirements in the EU . Retrieved from

European Union Agency for Fundamental Rights. (2018b). Consent for sexual activity with an adult . Retrieved from

Europol. (2017). Online sexual coercion and extortion as a form of crime affecting children—A Law Enforcement Perspective (pp. 1-23). European Cybercrime Centre.

Europol. (n.d.). Your life is online. Protect it! Retrieved from Europol website:

Facebook. (n.d.-a). Community conditions—Nude child and sexual exploitation of children. Retrieved from Https:// website:

Facebook. (n.d.-b). Community Standards Enforcement. Retrieved from Facebook Transparency website:

Ferenc, T. (2018). Nudity, Sexuality, Photography. Visual Redefinition of the Body. Qualitative Sociology Review , 14(2).

Finn, R. L., Wright, D., & Friedewald, M. (2013). Seven Types of Privacy. European Data Protection .

Frangež, D., Klančnik, A. T., Karer, M. Ž., Ludvigsen, B.-E., Kończyk, J., Pérez, F. R., ... Lewin, M. (2016). The Importance of Terminology Related to Child Sexual Exploitation 1. Journal of Criminal Investigation and Criminology , 66(4), 291-299

French Government. (2019). Creating a French Framework to make social media platforms more accountable: Acting in France with a European vision [Interim mission report].

Frosio, G. F. (2017). Reforming Intermediary Liability in the Platform Economy: A European Digital Single Market Strategy. Northwestern University Law Review Online , 112, 18-46

Gander, K. (2014, June 12). #FreeTheNipple: Facebook allows breastfeeding photos in policy change. Retrieved from The Independent website

Gillespie, A. A. (2011). Defining Child Pornography: Challenges for the Law (SSRN Scholarly Paper No. ID 1939870). Retrieved from Social Science Research Network website

Hart, M. (2017). Being naked on the internet: Young people's selfies as intimate edgework. Journal of Youth Studies , 20(3), 301-315.

Hasinoff, A. A. (2013). Sexting as media production: Rethinking social media and sexuality. New Media & Society , 15(4), 449-465.

Hasinoff, A., & Shepherd, T. (2014). Sexting in Context: Privacy Norms and Expectations. International Journal of Communication , 8, 2932-2415

Heins, M. (2013). The Brave New World of Social Media Censorship. Harvard Law Review Forum , 325-330

Instagram. (n.d.). Community Guidelines | Instagram Help Center. Retrieved from

Interagency Working Group on Sexual Exploitation of Children. (2016). ECPAT terminology guidelines—Luxembourg guidelines Sexual exploitation . Retrieved from

Internet Watch Foundation. (2015). Emerging Patterns and Trends Report #1 Online-Produced Sexual Content .

Ion, C., & Minea, C. (2019). Application of Image Classification for Fine-Grained Nudity Detection. In G. Bebis, R. Boyle, B. Parvin, D. Koracin, D. Ushizima, S. Chai, S. Sueda, X. Lin, A. Lu, D. Thalmann, C. Wang, & P. Xu (Eds.), Advances in Visual Computing (pp. 3-15). Springer International Publishing.

Jaeger, S. (2019, November 21). Commission takes stock of the fight against child sexual abuse and sexual exploitation of children . Migration and Home Affairs - European Commission. Retrieved from

Khan, U. (2008, December 30). Breastfeeding photo ban by Facebook sparks global protest by mothers. The Telegraph . Retrieved from

Kloess, J. A., Woodhams, J., Whittle, H., Grant, T., & Hamilton-Giachritsis, C. E. (2018). The Challenges of Identifying and Classifying Child Sexual Abuse Material. Sexual Abuse. A Journal of Research and Treatment , 1079063217724768.

Kobie, N. (2019, May 19). What is the dark web? How to use Tor to access the dark web. Wired UK . Retrieved from

Koelman, K. J., & Julia-Barcelo, R. (2000). Intermediary Liability in the E-commerce Directive: So Far, So Good, But It's Not Enough. Computer Law and Security Report , 4. Retrieved from

Koops, B.-J., Newell, B. C., Timan, T., Škorvánek, I., Chokrevski, T., & Galič, M. (2016). A Typology of Privacy (SSRN Scholarly Paper ID 2754043). Social Science Research Network.

Kuczerawy, A. (2018a). From 'Notice and Take Down' to 'Notice and Stay Down': Risks and Safeguards for Freedom of Expression (SSRN Scholarly Paper No. ID 3305153). Retrieved from Social Science Research Network website:

Kuczerawy, A. (2018b). Intermediary Liability and Freedom of Expression in the EU: From concepts to safeguards . Retrieved from Intersentia NV website

Kuczerawy, A. (2018c). Active vs. passive hosting in the EU intermediary liability regime: Time for a change? Retrieved from CITIP blog website

Kuczerawy, A. (2019). General Monitoring Obligations: A New Cornerstone of Internet Regulation in the EU? (SSRN Scholarly Paper No. ID 3449170). Retrieved from Social Science Research Network website

Kuczerawy, A., & Ausloos, J. (2015). From Notice-and-Takedown to Notice-and-Delist: Implementing Google Spain (SSRN Scholarly Paper No. ID 2669471). Retrieved from Social Science Research Network website

Kuczerawy, A., & Ombelet, P. J. (2015, July 2). Not so different after all? Reconciling Delfi vs. Estonia with EU rules on intermediary liability . Retrieved from CITIP blog website

Kuklis, L. (2019). Video-Sharing Platforms In AVMSD - A New Kind Of Content Regulation (SSRN Scholarly Paper ID 3527512). Social Science Research Network. Retrieved from

Lanzarote Committee. (2019a). Opinion of the Lanzarote Committee on child sexually suggestive or explicit images and/or videos generated, shared and received by children (No. T-ES(2019)02_en final)

Lanzarote Committee. (2019b). 'Sexting': Children who generate, receive and share sexual images of themselves do not produce and possess 'child pornography' [Press Release]. Retrieved from Council of Europe Communications website

Le Borgne-Bachschmidt, F., Girieud, S., Leiba, M., de Munck, S., Limonard, S., Poel, M., & Swart, E. (2008). User-created content: Supporting a participative information society: final report . Retrieved from

Lee, M., Crofts, T., Salter, M., Milivojevic, S., & McGovern, A. (2013, April 30). 'Let's Get Sexting': Risk, Power, Sex and Criminalisation in the Moral Domain. International Journal for Crime, Justice and Social Democracy .

Livingstone, S., & Mason, J. (2015). Sexual rights and sexual risks among youth online: A review of existing knowledge regarding children and young people's developing sexuality in relation to new media environments — LSE Research Online. Retrieved from

Livingstone, S., Davidson, J., & Bryce, J. (2017). Children's online activities, risks and safety: A literature review by the UKCCIS Evidence Group . Retrieved from UKCCIS website

Livingstone, S., Mascheroni, G., & Staksrud, E. (2018). European research on children's internet use: Assessing the past and anticipating the future. New Media & Society , 20(3), 1103-1122.

Livingstone, S., Ólafsson, K., & Staksrud, E. (2013). Risky Social Networking Practices Among 'Underage' Users: Lessons for Evidence-Based Policy. Journal of Computer-Mediated Communication , 18(3), 303-320.

Locatelli, E. (2017). Images of Breastfeeding on Instagram: Self-Representation, Publicness, and Privacy Management. Social Media + Society , 3(2), 2056305117707190.

Madigan, S., Ly, A., Rash, C. L., Van Ouytsel, J., & Temple, J. R. (2018). Prevalence of Multiple Forms of Sexting Behavior Among Youth: A Systematic Review and Meta-analysis. JAMA Pediatrics , 172(4), 327-335.

Mayer, F., Arent, T., Geserick, G., Grundmann, C., Lockemann, U., Riepert, T., Schmeling, A., Ritz-Timme, S., & Study Group on Forensic Age Estimation (AGFAD) of the German Association of Legal Medicine. (2014). Age estimation based on pictures and videos presumably showing child or youth pornography. International Journal of Legal Medicine , 128(4), 649-652.

McIntyre, T. J. (2011). Child Abuse Images and Cleanfeeds: Assessing Internet Blocking Systems (SSRN Scholarly Paper No. ID 1893667). Retrieved from Social Science Research Network website

Montagnani, M. L. (2019). A New Liability Regime for Illegal Content in the Digital Single Market Strategy (SSRN Scholarly Paper No. ID 3398160). Retrieved from Social Science Research Network website

Moreau, E. (2019, September 18). Hottest Social App Trends for Teens . Retrieved from Lifewire website:

Myers West, S. (2018). Censored, suspended, shadowbanned: User interpretations of content moderation on social media platforms. New Media & Society , 20(11), 4366-4383.

Nair, A. (2019). The Regulation of Internet Pornography: Issues and Challenges . Routledge, Abingdon, Oxon; New York, NY.

Nissenbaum, H. (2011). A Contextual Approach to Privacy Online. American Academy of Arts & Sciences , 32-48.

O' Neill, B. (2018). Research for CULT Committee - Child Safety Online: Definition of the Problem, European Parliament, Policy Department for Department for Structural and Cohesion Policies . Retrieved from

Ofcom. (2019). Children and parents: Media use and attitudes report 2018 . London

Ouvrein, G., & Verswijvel, K. (2019). Sharenting: Parental adoration or public humiliation? A focus group study on adolescents' experiences with sharenting against the background of their own impression management. Children and Youth Services Review , 99, 319-327.

Paul, K. (2019, June 3). Naked protesters condemn nipple censorship at Facebook headquarters. The Guardian . Retrieved from

Piwek, L., & Joinson, A. (2016). ``What do they snapchat about?'' Patterns of use in time-limited instant messaging service. Elsevier Enhanced Reader, 54, 358-367.

Quayle, E., & Cariola, L. (2019). Management of non-consensually shared youth-produced sexual images: A Delphi study with adolescents as experts. Child Abuse & Neglect , 95, 104064.

Ringrose, J., Harvey, L., Gill, R., & Livingstone, S. (2013). Teen girls, sexual double standards and 'sexting': Gendered value in digital image exchange. Feminist Theory , 14(3), 305-323

Sampathkumar, M. (2017, October 6). Facebook bans woman who shared article on breastfeeding. Retrieved from The Independent website

Sartor, G. (2017). Providers liability: From the eCommerce Directive to the future - In-depth analysis for the IMCO Committee, European Parliament.

Shayan, J., Abdullah, S. M., & Karamizadeh, S. (2015). An overview of objectionable image detection. 2015 International Symposium on Technology Management and Emerging Technologies (ISTMET), 396-400.

Smahel, D., Machackova, H., Mascheroni, G., Dedkova, L., Staksrud, E., Olafsson, K., Livingstone, S., & Hasebrink, U. (2020). EU Kids Online 2020—Survey results from 20 countries . Retrieved from

Snap. (n.d.). Community Guidelines - Snap Inc. Retrieved 25 September 2019, from

Solove, D. (2007). The Future of Reputation: Gossip, Rumor, and Privacy on the Internet . Yale University Press. Retrieved from

Steinberg, S. (2017). Sharenting: Children's Privacy in the Age of Social Media. UF Law Faculty Publications . Retrieved from

Stephen, B. (2019, January 30). Porn: You know it when you see it, but can a computer? The Verge . Retrieved from

Stewart, D. (2017). Social Media and the Law: A Guidebook for Communication Students and Professionals . Routledge

Taddeo, M., & Floridi, L. (Eds.). (2017). The Responsibilities of Online Service Providers . Law, Governance and Technology Series, Springer. Retrieved from

Taylor, M., & Quayle, E. (2003). Child pornography: An internet crime . Brunner/Routledge, Hove. Retrieved from

Tariq, M. U., Razi, A., Badillo-Urquiola, K., & Wisniewski, P. (2019). A Review of the Gaps and Opportunities of Nudity and Skin Detection Algorithmic Research for the Purpose of Combating Adolescent Sexting Behaviors. In M. Kurosu (Ed.), Human-Computer Interaction. Design Practice in Contemporary Societies (pp. 90-108). Springer International Publishing.

Timan, T., Newell, B. C., & Koops, B. J. (2017). Privacy in Public Space: Conceptual and Regulatory Challenges (No. 1). Edward Elgar Publishing. Retrieved from

Tolman, D. L. (2012). Female Adolescents, Sexual Empowerment and Desire: A Missing Discourse of Gender Inequity. Sex Roles , 66(11), 746-757.

Ullrich, C. (2017). Standards for Duty of Care? Debating Intermediary Liability from a Sectoral Perspective. JIPITEC 111, 8 .

van der Zwaan, J. M., Dignum, V., Jonker, C. M., & van der Hof, S. (2014). On Technology Against Cyberbullying. In J. van den Hoven, N. Doorn, T. Swierstra, B.-J. Koops, & H. Romijn (Eds.), Responsible Innovation 1: Innovative Solutions for Global Issues (pp. 369-392). Springer Netherlands.

Van Eecke, P., & Truyens, M. (2014). EU study on the legal analysis of a single market for the information society: New rules for a new age? Retrieved from

Van Eecke, Patrick. (2011). Online service providers and liability: A plea for a balanced approach. Common Market Law Review , 48 (5), 1455-1502

van Hoboken, J. (2012). Legal Space for Innovative Ordering: On the Need to Update Selection Intermediary Liability in the EU (SSRN Scholarly Paper No. ID 2013223). Retrieved from Social Science Research Network website

Van Royen, K., Poels, K., Daelemans, W., & Vandebosch, H. (2015). Automatic monitoring of cyberbullying on social networking sites: From technological feasibility to desirability. Telematics and Informatics , 32(1), 89-97.

Van Royen, K., Poels, K., & Vandebosch, H. (2016). Harmonizing freedom and protection: Adolescents' voices on automatic monitoring of social networking sites. Children and Youth Services Review , 64, 35-41.

Van Ouytsel, J., Van Gool, E., Walrave, M., Ponnet, K., & Peeters, E. (2017). Sexting: Adolescents' perceptions of the applications used for, motives for, and consequences of sexting. Journal of Youth Studies , 20(4), 446-470.

Van Ouytsel, J., Walrave, M., & Ponnet, K. (2018a). A Nuanced Account: Why Do Individuals Engage in Sexting? In M. Walrave, J. Van Ouytsel, K. Ponnet, & J. R. Temple (Eds.), Sexting: Motives and risk in online sexual self-presentation (pp. 39-51).

Van Ouytsel, J., Walrave, M., & Ponnet, K. (2018). Adolescent Sexting Research: The Challenges Ahead. JAMA Pediatrics , 172(5), 405-406.

Van Ouytsel, J., Walrave, M., Ponnet, K.& Temple, J. R. (2018). Sexting. In The International Encyclopedia of Media Literacy (pp. 1-6).

Van Ouytsel, J., Walrave, M., & Ponnet, K. (2019). Sexting within adolescents' romantic relationships: How is it related to perceptions of love and verbal conflict? Computers in Human Behavior , 97, 216-221.

Verswijvel, K., Walrave, M., Hardies, K., & Heirman, W. (2019). Sharenting, is it a good or a bad thing? Understanding how adolescents think and feel about sharenting on social network sites. Children and Youth Services Review , 104, 104401.

von der Leyen, U. (2019). Political Guidelines for the next European Commission (2019-2024)—"A Union that strives for more, My agenda for Europe" (p. 24). Retrieved from

Walrave, M., Van Ouytsel, J., Ponnet, K., & Temple, J. (2018). Sexting—Motives and risk in online sexual self-presentation . Retrieved from

WeProtect Global Alliance. (2016). Preventing and Tackling Child Sexual Exploitation and Abuse (CSEA): A Model National Response . Retrieved from

Wolak, J., Finkelhor, D., Walsh, W., & Treitman, L. (2018). Sextortion of Minors: Characteristics and Dynamics. The Journal of Adolescent Health: Official Publication of the Society for Adolescent Medicine , 62(1), 72-79.

Yoo, C. S. (2012). The Dynamic Internet: How Technology, Users, and Businesses are Transforming the Network (SSRN Scholarly Paper No. ID 2160397). Retrieved from Social Science Research Network website

[1] Argyro (Iro) Chatzinikolaou, Research group 'Law & Technology', Ghent University.

[2] A similar possibility is afforded to the Member States of the Council of Europe through Article 20(3) second indent of the Lanzarote Convention.

[3] In fact, the European Commission drafted a report assessing the implementation of the measures referred to in Article 25 of the Directive (European Commission, 2016d).

[4] Facebook (as well as Google - YouTube) is based in Ireland, and so the company's activities fall within the Irish jurisdiction(Broadcasting Authority of Ireland, 2019, p. 12). Nonetheless, the companyoutlines its ongoing relationship with the National Center for Missing and Exploited Children (NCMEC) and notes its collaboration with a number of safety experts, NGOs and national law enforcement authorities around the world "to disrupt and prevent the sexual exploitation of children across online technologies".

[5] At a national level, the broadness of the definition of child sexual abuse material depends on the transposition of the Directive and the description set out in the respective provision of the Criminal Code.

[6] Such an assertion is based on the definition of 'child pornography' included in the Directive 2011/93. At a national level, whether nude or semi-nude imagery depicting children in the absence of a primarily sexual purpose constitutes illegal CSAM shall depend on the definition laid down in the national legal framework.

[7] Whether parents or other third persons hold a right to post imagery depicting children and which involves nudity or a sexual element falls outside the scope of this article and shall not be discussed in the present section.

[8] See European Court of Human Rights, Stubing v Germany para 55; Norris v Ireland para 38; Dudgeon v United Kingdom para 41.