Social media companies like Facebook, Twitter and Google's YouTube have stepped up both the speed and number of removals of hate speech on their platforms in response to EU pressure to do more to tackle the issue, results of an EU evaluation seen by Reuters.
Facebook, Twitter, YouTube and Microsoft last year signed up to a voluntary code of conduct to take action in Europe within 24 hours, following rising concerns about the proliferation of racist and xenophobic content on social media triggered by the refugee crisis and attacks in Western Europe.
This included removing or disabling access to the content if necessary, better co-operation with civil society organisations and the promotion of "counter-narratives" to hate speech.
Facebook assessed notifications of hateful content in less than 24 hours in 58 per cent of cases, up from 50 per cent in December, according to a report by the European Commission, the main results of which were seen by Reuters.
Twitter also sped up its dealing with notifications, reviewing 39 per cent of them in less than 24 hours, as opposed to 23.5 per cent in December, when the Commission first reviewed the companies' progress and warned them they were being too slow.
YouTube, on the other hand, slowed down, reviewing 42.6 per cent of notifications in less than 24 hours, down from 60.8 per cent in December, the results showed.
"IT companies have all been improving time and response to notifications on manifest illegal hate speech," EU Justice Commissioner Vera Jourova said at a meeting of the EU High Level Group on combating racism, xenophobia and other forms of intolerance on Wednesday.
"There are differences among the companies… but we can objectively say that all have improved."
All the companies significantly increased the number of removals. Overall, content was removed in 59.2 per cent of cases, more than double the rate in December which was 28.2 per cent.
The proliferation of hate speech on social media has increased pressure on the companies to remove the content swiftly as they face the prospect of legislation at both EU and national level.
Last week EU ministers approved plans to force social networks to take measures to block videos with hateful content while the German government approved a plan in April to fine companies up to €50mil (RM241.14mil) if they fail to remove hateful postings quickly.
The most common ground of hate speech the Commission identified was xenophobia, including hatred against migrants and refugees, together with anti-Muslim hatred, followed by ethnic origin.
The spread of fake news and racist content has taken on more urgency in Germany after the arrival of about one million migrants over the last two years.