On 22 June 2020, the European Commission released the results of its fifth evaluation of the 2016 Code of Conduct on countering illegal hate speech online.
Advertisement
What is the aim of the Code?
The European Commission launched the Code of Conduct (hereafter, the Code) in May 2016 together with four major IT companies (Facebook, Microsoft, Twitter and YouTube) in an effort to respond to the proliferation of racist and xenophobic hate speech online.
The aim of the Code is to make sure that requests to remove content are dealt quickly with. When companies receive a request to remove from their online platform content deemed to be illegal, they assess this request against their rules and community guidelines and, where necessary, national laws transposing EU law on combatting racism and xenophobia. The companies commit to reviewing the majority of these requests in less than 24 hours and to removing the content if necessary, while respecting the fundamental principle of freedom of speech. To this date, eight companies have adhered to the Code, notably Facebook, YouTube, Twitter, Microsoft, Instagram, Dailymotion, Snapchat and Jeuxvideo.com.
How does the Commission evaluate the implementation of the Code?
The Code is evaluated through a monitoring exercise by a network of civil society organisations located in different EU countries. Using a commonly agreed methodology, these organisations test how the IT companies apply the Code in practice. They do this by regularly sending the IT companies requests to remove content and recording how long it takes the IT companies to assess the request, what the outcomes are, and the feedback they receive from the companies.
What are the main takeaways of the fifth monitoring exercise?
- Swift and stable response to illegal hate speech notified by users to platforms
The results of the fifth monitoring exercise, which also includes Jeuxvideo.com and Dailymotion, show that about 90% of the notifications are assessed within 24 hours. The IT companies fully meet the target of reviewing the majority of notifications within 24 hours. Facebook has reached 95.7% of notifications assessed within a day.
On average, IT companies remove 71% of illegal hate speech incidents notified to them by the NGOs and public bodies participating in the evaluation. Between 70% and 80% is estimated to be satisfactory removal rates, as some of the content flagged by users could relate to content that is not illegal. In order to protect freedom of speech only illegal content should be removed.
- Consistent and scrupulous assessment of illegal hate speech content
The monitoring exercise shows that the removal of content for serious cases of deemed illegal hate speech is very significant. Content that calls for murder or violent acts against certain groups is largely removed in most of the identified cases. Whereas, content using degrading, defamatory words or pictures to name certain social groups or individuals belonging to such groups are still removed in smaller proportions. This suggests that the review made by the companies is done with due consideration of protected speech.
- More efforts needed on transparency and feedback to users as well as to maintain high levels on response to notices, on a daily basis.
There are still some gaps in the information provided to users on the outcome of their notifications: while the average of notifications which received feedback is slightly higher than last year (67.1% vs. 65.4%), Facebook is the only platform that provides systematic feedback to almost all users. Instagram is around the average (61.5%) while the other platforms performs less well (Twitter 43.8%, Jeuxvideo.com 22.5%, and YouTube 8.8%).
An EU funded project carried out an additional evaluation adopting a very similar methodology, in a period of time outside the fifth monitoring exercise. Its results show that some IT companies had (slightly to significantly) lower removal rates compared to the results in the fifth monitoring exercise. The assessment of notices, and removal of illegal content when necessary, must be prompt and consistent over time.
How does the Code contribute to the wider work of the Commission on illegal content?
The results of the Code monitoring feed into the wider work of the Commission on the role of online platforms in the prevention, detection and removal of illegal content. The Commission is currently assessing how to address illegal hate speech online in the context of the Digital Services Act package, as part of a comprehensive regulatory framework on the responsibilities of information society services in address all forms of illegal content online. As a first step in the consultation process, the Commission has recently published a public consultation aimed at gathering views on the current issues around online platforms, including as regards their responsibilities to tackle illegal content online.
On 28 September 2018, the Commission adopted a Communication,which provides for guidance to platforms on notice-and-action procedures to tackle illegal content online. The importance of countering illegal hate speech online and the need to continue working with the implementation of the Code is reflected in this guidance document.
A Commission Recommendationon measures to tackle effectively illegal content online was published on 1 March 2018. It contains two parts, a general part on measures applicable to all types of illegal content and a specific part addressing the special actions that platforms would need to take to address terrorist content. In terms of the rules applicable to all types of illegal content the recommendation includes clearer ‘notice and action’ procedures, more efficient tools and proactive technologies, stronger safeguards to ensure fundamental rights, special attention to small companies and closer cooperation with authorities.
What is the definition of illegal hate speech?
Illegal hate speech is defined in EU law under the Framework Decision on combating certain forms and expressions of racism and xenophobia by means of criminal lawas the public incitement to violence or hatred directed to groups or individuals on the basis of certain characteristics, including race, colour, religion, descent and national or ethnic origin.
Is the Code the right solution to tackle hate speech online?
The self-regulatory approach set by the Code has proven to be an effective policy tool to achieve fast progress by the businesses in facing a major societal challenge. One policy alone cannot be the only solution to tackle the proliferation of hatred online. The Code focuses primarily of notice-and-action and removals and thus helps to treat the “symptoms”.
The challenges posed by hate speech online need to be tackled in a comprehensive way and the Commission intends to do so also in the context of the Digital Services Act package. The dissemination of illegal content online, including illegal hate speech, intermediated via online platforms creates an unsafe online environment for users across the EU. With the Digital Services Act, the Commission intends to create a horizontal regulatory framework for information society services on their obligations to address illegal content on their services.
How does the Commission work with the different IT platforms?
The Code is based on cooperation involving the European Commission, IT platforms, civil society organisations and national authorities. All stakeholders meet regularly under the umbrella of the High Level Group on combatting racism and xenophobia, to discuss challenges and progress. In addition to the regular monitoring exercises, the Commission engages in a constant dialogue with the platforms to encourage progress on all the commitments in the Code.
Workshops and trainings are also organised with companies and other relevant stakeholders. For instance, a workshop was held jointly with Google in Dublin in November 2017 to foster increased quality of notices by trusted flaggers and a more effective response by the companies’ content reviewers. This workshop was followed up by similar events co-organised with Facebook and Twitter in June 2018 and January 2019 respectively. A seminar on digital campaigns for counter narratives against hatred took place in Paris in December 2019.
Does the Code lead to censorship?
No. The Code aims to tackle online hate speech that is already illegal. The same rules apply both online and offline. In the Code, both the IT companies and the European Commission also stress the need to defend the right to freedom of expression. The Code cannot be used to make IT companies take down content that does not count as illegal hate speech, or any type of speech that is protected by the right to freedom of expression set out in the EU Charter of Fundamental Rights.
In addition, the results of a 2016 Eurobarometer surveyshowed 75% of those following or participating in online debates had come across episodes of abuse, threat or hate speech aimed at journalists. Nearly half of these people said that this deterred them engaging in online discussions. These results show that illegal hate speech should be effectively removed from social media, as it might limit the right to freedom of expression.
Is it not for courts to decide what is illegal?
Yes, interpreting the law is and remains the responsibility of national courts.
At the same time, IT companies have to act in line with national laws, in particular those transposing the Framework Decision on combatting racism and xenophobia and the 2000 e-commerce Directive. When they receive a valid alert about content allegedly containing illegal hate speech, IT companies have to assess it, not only against their rules and community guidelines, but, where necessary, against applicable national law (including that implementing EU law), which fully complies with the principle of freedom of expression.
Do all expressions of hatred qualify as illegal hate speech, e.g. “I hate you”?
Offensive or controversial statements or content might be legal. As the European Court of Human Rights said, “freedom of expression … is applicable not only to “information” or “ideas” that are favourably received or regarded as inoffensive or as a matter of indifference, but also to those that offend, shock or disturb the State or any sector of the population“.
In the Code, both the IT companies and the European Commission also stress the need to defend the right to freedom of expression.
Assessing what could be illegal hate speech includes taking into account criteria such as the purpose and context of the expression. The expression ‘I hate you’ would not appear to qualify as illegal hate speech, unless combined with other statements about for example threat of violence and referring to race, colour, religion, descent and national or ethnic origin, among others.
How can we prevent governments from abusing the Code?
The Code is a voluntary commitment made by the IT companies that have signed up to it. It is not a legal document and does not give governments the right to take down content. The Code cannot be used to make these IT companies take down content that does not count as illegal hate speech, or any type of speech that is protected by the right to freedom of expression set out in the EU Charter of Fundamental Rights.
Source: European Commission
