Armingaud Avocat

Commission Opinion of 20 January 2025 on the assessment of the Code of conduct on countering illegal hate speech online+ within the meaning of Article 45 of Regulation 2022-2065

31 min de lecture

Télécharger .md

[Access official publication]

COMMISSION OPINION of 20.1.2025 on the assessment of the Code of conduct on countering illegal hate speech online + within the meaning of Article 45 - Codes of conduct of Regulation 2022/2065

1. LEGAL BASIS AND PROCEDURE

(1) The Code of conduct on countering illegal hate speech online, originally agreed in May 2016 (‘the 2016 Code of Conduct’) by the providers of Facebook, Twitter (currently X) and YouTube and by Microsoft1, is aimed at preventing and countering the spread of illegal hate speech online. An additional eight signatories joined the 2016 Code of Conduct in the years that followed. At the time of this Opinion, the 2016 Code of Conduct counts 12 signatories. To date, seven of them are providers of online platforms designated by the Commission as very large online platforms (‘VLOPs’): Facebook, Instagram, LinkedIn, Snapchat, TikTok, X, YouTube. Other signatories are the providers of Dailymotion, Twitch, Rakuten Viber, Jeuxvideo.com and Microsoft.

(2) On 29 April 2024, the signatories to the 2016 Code of Conduct proposed a renewed set of commitments as part of a revised “Code of conduct on countering illegal hate speech online +” (hereafter “the Code”). In June 2024, those signatories informed the Commission of their willingness to have the Code assessed under 2065 of 19 October 2022 on a Single Market For Digital Services and amending Directive 2000/31/EC2.

(3) Pursuant to Article 45(1) of Regulation (EU) 2022/2065, “the Commission and the European Board for Digital Services [hereinafter ‘the Board’] shall encourage and facilitate the drawing up of voluntary codes of conduct at Union level to contribute to the proper application of this Regulation, taking into account in particular the specific challenges of tackling different types of illegal content and systemic risks […]“. Article 45(3) of Regulation (EU) 2022/2065 provides that: “[t]he Commission and the Board shall aim to ensure that the codes of conduct clearly set out their specific objectives, contain key performance indicators to measure the achievement of those objectives and take due account of the needs and interests of all interested parties, and in particular citizens, at Union level”. Under Article 45(4) of Regulation (EU) 2022/2065 “[t]he Commission and the Board shall assess whether the codes of conduct meet the aims specified in Article 45 paragraphs 1 and 3”. This Opinion sets out the assessment by the Commission. The assessment made by the Board is published separately”.

2. BACKGROUND

(4) Codes of conduct play an important role in the wider system of enforcement of Regulation (EU) 2022/2065, contributing to its application. While they remain voluntary, providers of VLOPs and of very large online search engines (‘VLOSEs’) may adhere to commitments made in such Codes as a means of fulfilling their obligation to put in place reasonable, proportionate and effective mitigation measures, pursuant to Article 35 of Regulation (EU) 2022/2065. This is without prejudice to the assessment by the Commission as to the appropriateness of such commitments to mitigate the risks identified under Article 34 which will be carried out on a case-by-case basis.

(5) Under Council Framework Decision 2008/913/JHA3, Member States must ensure that public incitement to violence or hatred directed against a group of persons or a member of such a group defined by reference to race, colour, religion, descent or national or ethnic origin (hereafter: ‘illegal hate speech’) is punishable under their national legislation. This Framework Decision also applies to public incitement to violence or hatred manifesting itself online. The scope of the 2016 Code of conduct makes explicit reference to content deemed illegal as per national laws transposing the Framework Decision. In addition, according to Directive 2024/1385 on combating violence against women and domestic violence, Member States shall ensure that intentionally inciting violence or hatred directed against a group of persons or a member of such a group, defined by reference to gender, by publicly disseminating, by means of ICT, material containing such incitement is punishable as a criminal offence.4

(6) While Regulation (EU) 2022/2065 provides a robust legal framework for countering illegal hate speech online, some of the voluntary commitments contained in the 2016 Code of Conduct have inspired new obligations under the Regulation (EU) 2022/2065 (see point (23) below). The 2016 Code of Conduct is cited in Regulation (EU) 2022/2065 in relation to processing notifications for removal of illegal hate speech online, and is listed as one of the existing self-regulatory instruments.

(7) The Code takes the form of five commitments and two annexes. Commitment 1 calls for clear terms and conditions that inform users about the policies and measures to address illegal hate speech online. Under Commitment 2 each signatory undertakes to review within 24 hours the majority (at least 50%) of hate speech notices received under Articles 16 and 22 of Regulation (EU) 2022/2065 from so-called ‘monitoring reporters’.5 According to commitment 3, each signatory undertakes to assess the adherence to the specific commitments on the review of notices and to monitor trends over time, following the methodology in Annex 1, and to provide additional information accompanying the results of each monitoring exercise referred to in Annex 2. Under commitment 4, signatories agree to participate in a multi-stakeholder framework for cooperation and information exchange (involving in particular the signatories and civil society organisations) and to participate in regular exchange fora. Under commitment 5, signatories agree to continue supporting tools and approaches on counter- and alternative narratives, new ideas and initiatives as well as educational programmes and to employ their best efforts to provide information about training, resources and support regarding best practices and how to use online platforms for delivering educational and awareness raising campaigns.

(8) Annex 1 outlines the methodology for the annual monitoring exercise under the Code, whereby monitoring reporters report alleged illegal hate speech online content through designated reporting processes for a period of maximum six weeks (30 working days). Detailed figures about the data points are published in a report within three months following the end of the monitoring period. The targets are those found in commitment 2 of the Code.

(9) Annex 2 lists key performance indicators as well as additional qualitative and quantitative information accompanying the result of each monitoring exercise that signatories commit to report in line with commitment 3 of the Code. These follow the five main commitments and their sub-elements under the Code.

3. ASSESSMENT

3.1. Assessment criteria

(10) Pursuant to Article 45(4) of Regulation (EU) 2022/2065, the Commission’s assessment of a particular code of conduct aims at verifying whether the conditions set out in Article 45(1) and (3) of that Regulation are fulfilled, that is whether:

(a) the code contributes to the proper application of the Regulation, taking into account in particular the specific challenges of tackling different types of illegal content and systemic risks, in accordance with Union law in particular on competition and the protection of personal data (Article 45(1));

(b) the code clearly sets out its specific objectives (Article 45(3));

(c) the code contains key performance indicators to measure the achievement of those objectives (Article 45(3)); and

(d) the code takes due account of the needs and interests of all interested parties, and in particular citizens, at Union level (Article 45(3)).

(11) The fulfilment of the aforementioned conditions should ensure that codes of conduct (i) clearly define the nature of the public interest objectives being addressed; (ii) contain mechanisms for independent evaluation of the achievement of their objectives; and (iii) clearly define the role of relevant authorities. Recital 103 of Regulation (EU) 2022/2065 indicates that particular attention should be paid to avoiding negative effects on security, the protection of privacy and personal data, as well as to the prohibition on imposing general monitoring obligations on the providers of intermediary services. While the implementation of codes of conduct should be measurable and subject to public oversight, this should not impair the voluntary nature of such codes and the freedom of interested parties to decide whether to participate.

(12) The compliance with commitments undertaken pursuant to the codes of conduct referred to in Articles 45, by providers of VLOPs and VLOSEs is subject to the independent audit obligations set out in Article 37(1)(b) of Regulation (EU) 2022/2065 and in Commission Delegated Regulation (EU) 2024/4366. Compliance with code commitments is also monitored by VLOPs’ and VLOSEs’ compliance officers under Article 41(3)(f) of Regulation (EU) 2022/2065. Since such codes involve stakeholders, the Commission and the Board, they can provide all parties with an opportunity to engage in a critical discussion of evidence and a forum for continuous engagement that complements Regulation (EU) 2022/2065 and regulatory enforcement. For this reason adherence to and compliance with a given code of conduct by a provider of VLOP or VLOSE may be considered as an appropriate risk mitigating measure under Article 35 of Regulation (EU) 2022/2065 (Article 35(1)(h)). However, it is important to note that the mere fact of participating in and implementing a given code of conduct should not in itself be a basis to presume compliance with the Regulation (EU) 2022/2065. As noted in Preamble (g) of the Code, DSA obligations take legal precedence over voluntary code commitments. This opinion does not prejudge any future action that the European Commission, the Board, any of the Digital Services Coordinators, and other competent authorities, as applicable, may undertake in the enforcement of Regulation (EU) 2022/2065.

3.2. Contribution to the proper application of the Regulation

(13) According to Article 45(1) of Regulation (EU) 2022/2065, the goal of a code of conduct should be “to contribute to the proper application of this Regulation, taking into account in particular the specific challenges of tackling different types of illegal content and systemic risks, in accordance with Union law, in particular on competition and the protection of personal data”.

(14) The Code aims to contribute to the proper application of Regulation (EU) 2022/2065 in relation to illegal hate speech online. It pursues the achievement of this aim through commitments intended to mitigate systemic risks associated with the dissemination of illegal content (Article 34(1)(a)), such as illegal hate speech online; actual or foreseeable negative effects for the exercise of fundamental rights (Article 34(1)(b), including the freedom of expression); actual or foreseeable negative effects on civic discourse and electoral processes, and public security (Article 34(1)(c)); and actual or foreseeable negative effects in relation to gender-based violence, the protection of public health and minors and serious negative consequences to the person’s physical and mental well-being (Article 34(1)(d)) stemming from concerns relating to the design, functioning or use of VLOPs and VLOSEs.

(15) Regulation (EU) 2022/2065 recalls that ‘illegal content’ is a concept that should be broadly defined, and that it should be understood to refer in particular to information, irrespective of its form, that the applicable rules render illegal in view of the fact that it relates to illegal activities, or that under the applicable law is itself illegal, such as illegal hate speech online. The definition employed by the Code (see footnote 4 of the Code) is aligned with these concepts, and it confirms that commitments set out in the Code are explicitly meant to mitigate the dissemination of illegal content online as a systemic risk.

(16) Additionally, the Code aims to mitigate risks associated with negative effects on the exercise of fundamental rights (Article 34(1)(b)). More specifically, its objective is to counter illegal hate speech online while protecting the right to freedom of expression and information. Values such as respect for “human dignity, freedom, democracy, equality, the rule of law and respect for human rights, including the rights of persons belonging to minorities” are enshrined in the Treaty on European Union7. All forms of hatred and intolerance are incompatible with these fundamental rights and values. Section 2 of the Code in particular seeks to address the need to strike a careful balance between measures to remove or disable access to alleged online hate speech content and the protection of the freedom of expression and information. This Section includes a specific target as regards the review time of hate speech notices, setting it to 24 hours for the majority of notices and taking into account that the time necessary to action a notice may depend on contextual factors that complicate the assessment of illegality.

(17) The Code’s engagement with protecting the freedom of expression and information aims to have further effects. Hate speech can have profound consequences for the victims, and impact negatively their mental and physical health and social and economic wellbeing. It may also target whole groups and communities and the persons forming these groups and communities by rejecting and devaluing them. Hate speech can contribute to radicalisation, violent extremism and terrorism. It produces a silencing effect on individuals belonging to groups at risk of hate victimisation, thus hampering the right to freedom of expression. Hate speech not only wounds the individuals, groups and communities directly targeted by it, but it also generates polarisation and inhibits a free and vibrant democratic debate. Manifestations of hatred, including online hate speech, undermine society as a whole and threaten the foundations of our democracies. The Code is precisely aimed at tackling illegal hate speech online, thus contributing to the public interest objectives of preventing and combatting crime, protecting freedom of expression and information, increasing security by preventing radicalisation, violent extremist or terrorist ideologies, fighting racism and xenophobia, and protecting victims of crimes. In this way the Code also seeks to combat negative effects on civic discourse and electoral processes, and on public security (Article 34(1)(c) of Regulation (EU) 2022/2065).

(18) The Commission also takes note of the breadth of the notion of illegal hate speech that the Code embraces. Footnote 4 of the Code indicates that it applies to illegal hate speech “as defined by applicable laws, including the Framework Decision 2008/913/JHA of 28 November 2008 on combating certain forms and expressions of racism and xenophobia by means of criminal law, as transposed in national jurisdictions, as well as possible forthcoming updates to this Framework Decision, where relevant”. This means that all conducts defined as hate speech, both in the laws transposing the Framework Decision and in any other provisions of national law, constitute hate speech for the purposes of the Code and should be reported and removed accordingly. Updates in the EU legal framework may derive e.g. from the adoption of a Council Decision extending to hate speech and hate crimes the current list of ‘EU crimes’ as laid down in Article 83(1) of the Treaty on the Functioning of the European Union (‘TFEU’) and any ensuing EU legislation enlarging the current list of protected grounds under the Framework Decision8. Legal developments may also concern the criminalisation of specific forms of gender-based hate speech (misogyny online) as defined in Directive (EU) 2024/1385 on combating violence against women.

(19) In addition, the commitments set out in the Code can also contribute to the mitigation of risks related to gender-based violence or hatred and risks for children’s mental health and well-being (Article 34(1)(d)). This is notably in view of the fact that any intersectional manifestation of illegal hate speech (i.e. illegal hate speech with multiple bias motivations) also falls within its scope, and considering that the Code’s overarching objective to limit the spread of illegal hate speech on online platforms can help to reduce the potential exposure of children to illegal hate speech online.

(20) In Section 4 of the Code, the signatories commit to enhancing intra-industry and multi-stakeholder cooperation to help prevent risks of the spread of illegal hate speech online as well as the understanding of the specificities of illegal hate speech online in different national/linguistic contexts. In particular, the signatories commit to participate in a regular and structured exchange on best practices, trends and developments related to illegal hate speech online (exchange fora). In parallel, the establishment of a shared online knowledge hub aims to enhance best practice resources, including relevant information on the challenges and opportunities presented by emerging technologies in both the generation and detection of illegal hate speech online. Section 4 of the Code therefore aims to mitigate the risks related to the spread of illegal hate speech online through early warning, exchange of knowledge and commitments for quick response in the event of a material risk of a significant increase in the dissemination of illegal hate speech on the signatories’ platforms. Section 5 aims to raise general awareness and promote education initiatives to foster civility online and prevent the proliferation of illegal hate speech that can contribute to societal polarisation and other risks deriving from the proliferation of hatred and intolerance.

(21) Article 45(1) of Regulation (EU) 2022/2065 further requires codes of conduct to be drawn up in accordance with Union law, in particular on competition and protection of personal data. Regarding data protection, it is notable that Point 4.4 of the Code requires that any intra and cross-platform exchanges about operations and relevant trends respect privacy legislation and take into due consideration threats to fundamental rights. The fact that signatories commit to promoting the respect of fundamental rights as enshrined in the Charter of Fundamental Rights of the EU should contribute to ensuring that the right to personal data protection is safeguarded in the interpretation and application of the Code. Additionally, an assessment of the Code’s scope and commitments does not, in principle, appear to raise material concerns of a competition law nature, to the extent that it does not involve the disclosure of commercially sensitive information among the signatories. Moreover, the Code does not exclude new signatories from participating in the future, and sets out transparent measures that can be replicated by other industry players to ensure the safety of their users and the protection of their freedom of expression.

(22) The Code acknowledges that the actions it generates must build upon and complement the regulatory requirements and overall objectives of Regulation (EU) 2022/2065, as evidenced in Preamble (e). In particular, Preamble (f) recognises that the Code is meant to create “a framework that facilitates the compliance with and the effective enforcement of Regulation (EU) 2022/2065 in the specific area of illegal hate speech content”.

(23) In that regard, certain commitments (such as for example in Points 1.1, 1.2, 2.1 and 2.2) under the Code are drafted in a language that mirrors that of certain provisions of Regulation (EU) 2022/2065. Each of the references to Regulation (EU) 2022/2065 aims to acknowledge that the Code represents a further operational and complementary layer on top of existing legal obligations in the specific focus area of countering illegal hate speech online. Additionally, the Code recognises that it is without prejudice to the obligations imposed on the signatories under Regulation (EU) 2022/2065, where applicable (Preamble (g)), and that it aims to gather “specific and proportionate information on how obligations are implemented with respect to illegal hate speech” (Preamble (h)). In this way, the Code has the benefit of allowing additional information to be gathered through signatories’ reports since information on complementary Code-based commitments is given in addition to what signatories are already required to report under Regulation (EU) 2022/2065. (24) The commitments in the Code are complementary to Regulation (EU) 2022/2065. Point 2.3, for example, provides a review time baseline that complements the obligations found in Articles 16 and 22 of Regulation (EU) 2022/2065. The methodology for the monitoring exercises and the system of ‘Monitoring Reporters’ set out in Annex 1 complement Regulation (EU) 2022/2065 and establish a system for independent monitoring by those reporters on Point 2.3 of the Code, without prejudice to the system of ‘Trusted Flaggers’ provided for in Article 22 of Regulation (EU) 2022/2065.

(25) Annex 2 on reporting (i.e. reporting by the signatories on the measures taken to address illegal hate speech online as part of their content moderation policies), specifies, in point d), that compliance and complementarity with the minimum requirements set out in Regulation (EU) 2022/2065 is achieved by setting the reporting periods of 1 January - 30 June and of 1 July - 31 December. This can contribute to streamlining and thereby reducing the cost of reporting obligations.

(26) In light of the above, the Commission concludes that the Code contributes to the proper application of the Regulation. It complements and builds upon the regulatory requirements of Regulation (EU) 2022/2065 and does not appear to contain, in the Commission’s assessment, any incompatibilities with EU law.

(27) However, the Commission would like to stress a number of points to be taken into account in the implementation of the Code when it comes to its contribution to the proper application of the Regulation. These points should inform the monitoring and evaluation of the Code and any possible future review and adaptation thereof by the signatories pursuant to Article 45(4) of Regulation (EU) 2022/2065.

(28) In the context of reporting under point 2 of Annex 2 about the commitment 2.2 of the Code and without prejudice to the Commission’s assessment of signatories’ compliance with Regulation (EU) 2022/2065 in individual cases, the Commission encourages signatories to make a distinction between the grounds (i.e. terms and conditions or obligations stemming from EU law and/or national laws) on which content moderation decisions are taken. It is important to underline that signatories have different transparency and user notification obligations under the DSA depending on whether removals are made on the grounds of their terms of service or on the basis of the applicable law.

(29) Regarding Annex 1 reporting (i.e. reporting in the context of the annual monitoring exercise), the Commission encourages signatories to achieve additional complementarity by aligning the timing of the publication of the results of the monitoring exercise with the reporting under Annex 2 and transparency reporting under Regulation (EU) 2022/20659.

3.3. Whether specific objectives are clearly set out

(30) According to Article 45(3) of Regulation (EU) 2022/2065, the Commission should aim to ensure that codes of conduct clearly set out their specific objectives.

(31) With reference to the Code’s objectives described in paragraphs (14), (16) and (20) the Code establishes in commitment 2 a clear benchmark in relation to the maximum time of review of a majority (at least 50%) of notices submitted by monitoring reporters (24 hours). Moreover, signatories commit their best efforts to reach a target of at least 67% of notices reviewed within that time frame. Annex 1 complements this particular commitment with a detailed methodology to monitor these objectives in a way that is predictable, transparent and recognised by all involved parties (i.e. monitoring reporters and providers of online platforms).

(32) As part of Section 3 and Annex 2, the Code contains specific reporting objectives to share additional qualitative and quantitative information about the signatories’ content moderation in relation to illegal hate speech online10. In particular, Annex 2 requires that providers of platforms report on measures taken to provide training and assistance to persons in charge of content moderation, as well as measures taken that affect the availability, visibility and accessibility of information about illegal hate speech content online and the trends detected.

(33) In addition, as part of its Section 4, the Code identifies clear deliverables, e.g. exchange fora, an annual meeting, a structured dialogue and information exchange through a standard set of questions, an online knowledge hub and ad hoc dialogue between signatories and the Commission in the event of a material risk of a significant increase in the dissemination of illegal hate speech on the signatories’ online platforms.

(34) In Section 5, the Code aims to reinforce cooperation between civil society organisations and platforms on actions to counter hatred online, such as awareness raising and educational initiatives to promote civility and safety online. It also aims to establish clear commitments to enhance training and capacity building on how to make best use of the platforms’ services in this respect.

(35) In light of the above, the Commission concludes that the Code clearly sets out its specific objectives.

3.4. Whether the Code contains key performance indicators

(36) Article 45(3) of Regulation (EU) 2022/2065 stipulates that codes of conduct must contain key performance indicators (‘KPIs’) to measure the achievement of their specific objectives.

(37) The Code contains KPIs to measure its specific objectives. Annex 2 to the Code contains a table which summarises how each commitment set out in Sections 1, 2, 3, 4 and 5 of the Code will be monitored.

(38) In Section 3, the signatories commit to assess the adherence to the specific commitments on the review of notices (under Point 2.3), to monitor trends over time and to provide additional information on content moderation in relation to hate speech online, accompanying the results of the monitoring exercises (Point 3.2).

(39) Annex 1 establishes a yearly monitoring exercise carried out by monitoring reporters, online. They take note of the reaction time and of the measures taken and include the results in a publicly available report. The yearly monitoring exercise ensures a thirdparty independent verification of the review time commitment and an approach to monitoring of KPIs that is not solely based on information shared by the signatories. The role of the monitoring reporters is primarily linked to the annual monitoring exercise described under Annex 1 to the Code. It is without prejudice to the system of trusted flaggers established under Article 22 of Regulation (EU) 2022/2065. At the same time, should entities who have been awarded the status of Trusted Flagger also become monitoring reporters, their role as Monitoring Reporters is without prejudice to the independence requirements laid down in Article 22(2) of the Regulation (EU) 2022/2065.

(40) The Commission intends to pay particular attention to the governance of the monitoring exercise in the context of the regular review of the Code foreseen under Article 45(4) of Regulation (EU) 2022/2065, so as to make sure that it provides a genuine added value to the framework of content moderation, including the function of Trusted Flaggers.

(41) Annex 2 lists KPIs to be used to gather quantitative and qualitative information on the commitments taken in relation to the moderation of hate speech content by the signatories. It sets out clearly identified parameters/indicators, e.g., amendments to the terms and conditions, systems for reporting, paths for submitting notices and their moderation, any proactive measure taken to detect, review and moderate hate speech on the platforms as well as relevant information on the role of recommender systems.

(42) The KPIs are fed largely by information provided by the signatories in accordance with Annex 2, with the addition of the results of the monitoring exercises detailed in Annex 1. Annex 2 also provides for a list of questions to assess the signatories’ commitment regarding the cooperation among industries and with stakeholders, by referring to, inter alia, information on the signatories’ participation and contribution to the regular exchange fora referred to in Section 4.2 of the Code, including the annual meeting, the intra-industry exchange of best practices, the structured dialogue between signatories and Monitoring Reporters, and the online knowledge hub. Some of the information will also be complemented by civil society organisations (e.g. as regards the support to Monitoring Reporters and any relevant civil society organisations in Point 5.2).

(43) The Code includes mechanisms for the independent evaluation of some of its key commitments, as outlined in paragraph (39) above.

(44) As regards Annex 2, the information, largely provided by signatories as detailed above, will be subject to public scrutiny. In addition, experts and civil society organisations could provide additional input and contribute to a collective and transparent assessment of the information provided. Such additional input may for example complement the information provided by signatories under Points 1(2) on the changes in their terms and conditions or Point 2(3) of Annex 2 on their reporting systems.

(45) In light of the above, the Commission concludes that the Code contains key performance indicators to measure the achievement of its objectives.

(46) However, the Commission would like to stress a number of points to be taken into account in the implementation of the Code as regards the KPIs. These points should inform the monitoring and evaluation of the Code and any possible future review and adaptation thereof by the signatories pursuant to Article 45(4) of Regulation (EU) 2022/2065.

(47) In line with Articles 45(2) and (3), and without prejudice to the Commission’s assessment of signatories’ compliance with Regulation (EU) 2022/2065 in individual cases, the Commission encourages signatories to report not only on measures taken, but also on the outcome of those measures when implementing the commitments of the Code. This could entail sharing metrics and data that measure the impact of content moderation decisions, e.g. by making use of the “Looking ahead” section (Points 13 and 14) of Annex 2, which specifically encourage the signatories to report additional information and learning in relation to the subject matter of the Code. As concerns data produced by the monitoring exercise under Annex 1, the Commission notes that it may be relevant to also extract information at the level of individual Member States (by platform, language or by categories of “grounds of hate” to which the content relates). Such patterns on a national level may greatly increase the analytical value of the Code.

(48) With a view to further strengthening the achievement of the objectives of the Code, and without prejudice to the Commission’s assessment of signatories’ compliance with Regulation (EU) 2022/2065 in individual cases, the Commission also encourages signatories to make use of the above mentioned Points 13 and 14 of Annex 2 to expand their reporting beyond the list of reference points under the previous sections of Annex 2, for example by including KPIs capturing the reach of illegal content prior to its removal, if possible distinguishing between organic and algorithmic reach, when implementing the commitments of this Code. Such data may also become useful in comparisons over time and across platforms, especially when viewed against the total number of actioned hate speech content per platform, which relevant signatories have to disclose as part of their transparency reporting obligations under Article 15 of Regulation (EU) 2022/2065.

(49) In the context of reporting under Annex 2, and without prejudice to the Commission’s assessment of signatories’ compliance with Regulation (EU) 2022/2065 in individual cases, the Commission also encourages signatories to provide more detailed information about the moderation path for hate speech notices when implementing the commitments made in the Code. An important element in that regard would be the inclusion of information about the internal classification of hate speech (for example, the grounds of hate speech such as race, ethnicity, religion, gender identity or sexual orientation), the training of classifiers for systems for the detection of hate speech in all official EU languages, the accuracy of classifiers (such as error rates), the number of appeals requested by users and the error rate of human and automatic content moderation decisions.

(50) Without prejudice to the Commission’s assessment of signatories’ compliance with Regulation (EU) 2022/2065 in individual cases, the Commission encourages the use of a more detailed research-friendly reporting template to facilitate consistent and comparable reporting of information provided under Annex 2.

(51) The aforementioned is also without prejudice to the mandatory monitoring and evaluation mechanisms for codes of conduct in Regulation (EU) 2022/2065. Under Article 37(1)(b), the providers of VLOPs and VLOSEs must undergo an independent audit, at their own expense and at least once a year. That audit will contain the assessment of their compliance with commitments undertaken pursuant to codes referred to in Article 45. Additionally, as already mentioned, under Article 41(3)(f), the providers of VLOPs and VLOSEs must establish a compliance function consisting of one or more compliance officers who must monitor compliance with commitments made pursuant to codes referred to in Article 45.

3.5. Whether due account is taken of the needs and interests of all interested parties, in particular citizens, at Union level

(52) Article 45(3) of Regulation (EU) 2022/2065 provides that the Commission should aim to ensure that codes of conduct take due account of the needs and interests of all interested parties, in particular citizens.

(53) As described in paragraphs (16) and (17) above as well as in Preamble (c), the fight against illegal hate speech online is first and foremost a way to promote and protect citizens’ fundamental rights to dignity, freedom of expression and nondiscrimination. The right to freedom of expression and information in particular is enshrined in Article 11 of the Charter and covers not only ‘information’ or ‘ideas’ that are regarded as inoffensive, but also those that offend, shock or disturb11. However, human rights law acknowledges that freedom of expression must not be exploited to incite hatred and violence12.

(54) The involvement of civil society organisations active in the Member States in monitoring the Code as well in the multi-stakeholder cooperation and initiatives on awareness raising to foster online civility enhances the representation of wider societal interests and the participation of diverse interest groups into the process of tackling hate speech online. A wide range of civil society organisations are involved prominently in the monitoring exercise described in Annex 1, as well as in the context of the exchange fora set out in Section 4 of the Code.

(55) In light of the above, the Commission concludes that the Code takes due account of the needs and interests of all interested parties, in particular citizens, at Union level.

(56) When it comes to implementation and possible future review, the Commission encourages that signatories to duly take into account the importance of multistakeholder cooperation to enhance the level of awareness of the most urgent societal threats posed by illegal hate speech, and a comprehensive exchange on most worrying trends as regards hate speech in specific national contexts or as affecting specific communities and groups.

4. CONCLUSION

(57) In light of the foregoing, the Commission concludes that the Code of conduct on countering illegal hate speech online + meets the conditions specified in Article 45(1) and (3) of Regulation (EU) 2022/2065.

(58) In accordance with Article 45(4) of Regulation (EU) 2022/2065, the Commission shall regularly monitor and evaluate the achievement of the objectives of the Code of conduct on countering illegal hate speech online + and encourage and facilitate the regular review and adaptation of that code.

Done at Brussels, 20.1.2025 For the Commission Henna Virkkunen Executive Vice-President

Footnotes

  1. Microsoft-hosted consumer services, as relevant. ↩

  2. OJ L 277, 27.10.2022, p. 1. ↩

  3. Council Framework Decision 2008/913/JHA of 28 November 2008 on combating certain forms and expressions of racism and xenophobia by means of criminal law, OJ L 328/55, 6.12.2008. See in particular Article 1(1)(a). ↩

  4. Directive (EU) 2024/1385 of the European Parliament and of the Council of 14 May 2024 on combating violence against women and domestic violence, OJ L 24.5.2024. See in particular Article 8. ↩

  5. Monitoring reporters are not-for-profit or public entities with expertise on illegal hate speech in at least one of the Member States who report content to the signatories that they consider, to the best of their knowledge, to be illegal hate speech online. Monitoring reporters can also be Trusted Flaggers as set out in Article 22 of the DSA, as relevant. ↩

  6. Commission Delegated Regulation 2024/436 of 20 October 2023 supplementing Regulation (EU) 2022/2065 of the European Parliament and of the Council, by laying down rules on the performance of audits for very large online platforms and very large online search engines (OJ L, 2.2.2024). ↩

  7. Article 2, Consolidated Version of the Treaty on European Union, OJ C 326/17, 26 October 2012. ↩

  8. See the Commission’s Communication COM/2021/777(final) on ‘A more inclusive and protective Europe: extending the list of EU crimes to hate speech and hate crime’ which aims to trigger a Council Decision extending to hate crime and hate speech the current list of so-called ‘EU crimes’ as laid down in Art 83(1) TFEU. ↩

  9. Annex 1, Introduction b) explicitly calls for a connection of these elements since it states that the results of each Signatory’s monitoring exercise have to be seen in combination with the additional information on the measures taken to address illegal hate speech based on the Signatory’s content moderation policies, proactive efforts, and enforcement actions provided according to the guiding questions set out in Annex 2. ↩

  10. As defined in paragraph (5) of this document, under Council Framework Decision 2008/913/JHA of 28 November 2008 (etc.) the Member States must have ensured that public incitement to violence or hatred directed against a group of persons or a member of such a group defined by reference to race, colour, religion, descent or national or ethnic origin is punishable under their national legislation (‘illegal hate speech’). ↩

  11. European Court of Human Rights, Handyside v United Kingdom, 7 December 1976, Series A no. 24, paragraph 49; European Court of Human Rights, Soulas and Others v France, 10 July 2008, 15948/03, paragraph 35. ↩

  12. European Court of Human Rights, Erbakan v Türkiye (6 July 2006, 59405/00, paragraph 56) whereby the ECtHR has confirmed that it “may be considered necessary in democratic societies to sanction or even prevent all forms of expression which spread, incite or justify hatred based on intolerance”. ↩

Vue Graphique