Hate Speech

The European approach to regulating and moderating the spread of hate speech online represents an interesting experiment in public-private collaboration.

A triad of normally unaffiliated institutions sit at the middle of this effort. Non-governmental organisations (NGOs) – particularly organisations with a background in promoting tolerance and fighting discrimination – are empowered to follow what happens on the platforms and "notify" content which they believe constitutes illegal hate speech. The platforms, in turn, evaluate 1) the notifications, 2) the local laws of the countries where a particular piece of content appears, and 3) their own community standards and guidelines. Based on that assessment, they decide if the content should be allowed to stay or taken down. Later the NGOs produce a report evaluating how much of the content they notified was taken down. The European Commission sits in the middle of the process – facilitating communication among the parties and providing periodic assessments of how much hate speech is slipping through despite the effort.

The unusual collaboration grew from equally unusual roots. Rather than imposing regulation, the European Commission encouraged platforms and NGOs to work together on this. Together, the three parties drew up a code of conduct on countering hate speech online (2016). To date, Dailymotion, Facebook, Google Plus, Instagram, Jeuxvideo, Microsoft, Snapchat, Twitter and YouTube have all signed on. Periodic assessments – fed by national NGO reports – have judged progress. You can find out more about this remarkable experiment on the European Commission code of conduct web page. Evidence from the monitoring exercise appears below as well.

The code itself draws on a definition laid down in the European Union’s framework decision on combating certain forms and expressions of racism and xenophobia by means of criminal law (2008). It defines hate speech as "all conduct publicly inciting 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, when carried out by the public dissemination or distribution of tracts, pictures or other material; Publicly condoning, denying or grossly trivialising crimes of genocide, crimes against humanity and war crimes […], when the conduct is carried out in a manner likely to incite violence or hatred against such a group or a member of such a group."

More information about hate speech and the effort to regulate it is available on the World Intermediary Liability Map (WILMap), led by the Center for Internet and Society at Stanford Law School.

Records 31 - 33 of 33


chart preview

Trend in Removal Rates on Twitter Based on the Moving Averages of Percentage of Removed Cases (2017)

The chart presents the share of the reported content which was removed by Twitter, based on data collected by the International Network Against Cyber Hate. The report found that, in 2017, Twitter's monthly removal rate has a high variation, and recorded a maximum level in February (90%) and a minimum in November (around 10%). Overall, Twitter’s removal rate shows a steep downward trend in 2017.
chart preview

Trend in Removal Rates on Youtube Based on the Moving Averages of Percentage of Removed Cases (2017)

The chart presents the share of the reported content which was removed by YouTube, based on data collected by the International Network Against Cyber Hate. The report found that, in 2017, YouTube’s monthly removal rate was highly volatile, recording maximum levels in January and April (90%) and minimum one in June (around 22%). Overall, YouTube’s removal rate has a slight downward trend in 2017.
chart preview

Trend of the Rate of Notifications Assessed Within 24 Hours

The chart shows the trend of the share of notifications reviewed within 24 hours by the tech companies over the period December 2016 - Octomber 2021, based on the data reported by social media platforms participating in the European Commission's Code of conduct. In 2021, Jeuxvideo.com assessed all notifications in less than 24h, while YouTube did so for 88.8% and Facebook for 81.5%. Among the platforms participating in the Code of conduct, Instagram has the lowest rate of review (62.4%). On average, the rate of review of companies remain high (81%), but shows a slight decline compared to 2020 (90.4%).