Illegal Content

The law is clear; what is illegal offline is illegal online. And a bevy of European laws – such as the directive on combating terrorism (2017) and the directive on combating the sexual abuse and sexual exploitation of children and child pornography (2011) – have been promulgated over the years, requiring platforms to remove illegal content "expeditiously," in the words of the electronic commerce directive (2000), once they are notified or in some other way become aware of its existence.

A short list of offences covered by these rules includes incitement to terrorism, xenophobic hate speech with intent to incite, copyright infringement and child abuse. Few dispute the underlying principles at stake, but the debate grows heated over how – and how quickly – platforms should be legally bound to remove illegal material when it appears. How much redress do platform users have if they believe their content was removed unjustifiably? And how much duty do the platforms have to cooperate pro-actively with security and law enforcement services?

The European Commission has provided a handy framework: a communication on tackling illegal content online (2017) and a recommendation on measures to effectively tackle illegal content online (2018). These guidelines are non-binding, but – faced with a rise in phenomena like terrorism – some governments have stepped in with steeper, more binding rules. Germany is a case in point. Its Netzwerkdurchsetzungsgesetz (NetzDG), threatens large fines if "manifestly unlawful" material does not disappear from platforms within 24 hours of notification. France’s proposed Loi contre les contenus haineux sur Internet (the "Avia law," named for sponsor Laetitia Avia), would do the same.

Put simply, the discussion boils down to several simple determinations: is illegal content coming down quickly enough? Are the rules and codes of conduct strong enough to prevent damage from occurring given the speed with which harm can take place? Is the volume of illegal content decreasing given the measures already in place, and is it decreasing quickly enough? And if stronger measures are needed, how can they be constructed to obtain better results without violating important rights such as privacy, redress and free speech?

The evidence presented in this section cover illegal content broadly. Separate sections will deal with more concrete aspects, such as incitement to terrorism, hate speech and copyright infringement. Additional information on illegal content online can be found on the World Intermediary Liability Map (WILMap), led by the Center for Internet and Society at Stanford Law School.

Records 41 - 50 of 64


chart preview

Percentage of Content Found by Facebook as Containing Child Nudity and Sexual Exploitation Compared to the Content Reported by the Users

This chart shows the percentage of content found by Facebook as containing child nudity and exploitation compared to the content reported by the users from July 2018 until March 2021. Since April 2021, the Child Nudity and Sexual Exploitation content have been renamed Child Endangerment, with two categories: Nudity and Physical Abuse and Sexual Exploitations, which are monitored separately. The data shows that the share of content reported by users (around 1%) is significantly lower that the one found by Facebook.
chart preview

Percentage of Content Found by Facebook as Containing Child Sexual Exploitation Compared to the Content Reported by the Users

This chart shows the percentage of content found by Facebook as containing child sexual exploitation compared to the content reported by the users from April 2021 until March 2022. The percentage reported by users is significantly lower that the one found by Facebook.
chart preview

Percentage of Content Found by Facebook as Containing Spam Compared to the Content Reported by the Users

This chart shows the percentage of content found by Facebook as containing spam compared to the content reported by the users over the period October 2017 - March 2022. As the results show, almost all of the content containing spam was first found by Facebook (the percentage remains above 99.5% for the whole period).
chart preview

Percentage of Content Found by Instagram as Containing Child Nudity and Exploitation Compared to the Content Reported by the Users

This chart shows the percentage of content found by Instagram as containing child nudity and exploitation compared to the content reported by the users, from April 2019 until March 2021. Since the April 2021, the child nudity and sexual exploitation content was renamed child endangerment and includes two separate categories - nudity and physical abuse and sexual exploitations, which are monitored separately. The data shows that the percentage of the content found by Instagram improved constantly over the years, reaching 98% in the third quarter of 2021.
chart preview

Percentage of Content Found by Instagram as Hate Speech Compared to the Content Reported by the Users

This chart shows the percentage of content found by Instagram as containing hate speech compared to the content reported by the users, over the period October 2019 - March 2022. Since April 2020, the hate speech detection of Instagram improved significantly compared to the beginning period. However, in the first quarter of 2022, the volume of content found by Instagram declined by 3.8% compared to the same period of the previous year.
chart preview

Percentage of European Businesses who Support Proposed Changes to the Single Market

This chart exhibits that an overwhelming majority of Europeans businesses support enacting reforms to the current single market structure in order to promote expansion and further development across sectors. The most favored reform was cutting redtape, including but not limited to "extensive reporting, information or documentation obligations."
chart preview

Products Affected by Government Requests of Removal

The chart presents the distribution of the most affected products by governments' requests of removal, since 2009. The shares are calculated based on the total numbers of requests received by Google from governments since July 2009 until December 2021. The data shows that the products with the most frequent government requests to remove content are YouTube, Web search and Blogger, but other products are also affected.
chart preview

Share of Enterprises Who Pay to Advertise Online, 2018

The chart presents the share of entreprises that pay for on-line advertising in European Union in 2018, by size of enterprise (the chart does not include the enterprises from the financial sector).
chart preview

Share of Fake Accounts and Spam Content Actioned on Facebook (2017-2022)

The chart shows the share of fake accounts and spam content actioned on Facebook, from the fourh quarter of 2017 until the first quarter of 2022. While these two violations remain the main reasons of removal of content on Facebook, the data shows that the other types of violations (such as adult nudity and sexual activity, child nudity and sexual exploitation, bullying and harassment, dangerous organisations, hate speech, and violent and graphic content) have also increased during this period.
chart preview

Share of Reported Content Removed by Google Under Germany’s Network Enforcement Act, by Reason of Removal (2018-2021)

The chart presents the share of the reported content removed or blocked by Google due to violation of the Germany’s Network Enforcement Act, over the period 2018-2021, by reason of removal. The data shows that hate speech or political extremism, privacy and terrorist or unconstitutional content have the highest share of removals, while violence has the lowest one.