Twitter and some of its social media competitors lagged behind this year in removing hate posts that are illegal in Europe, regulators said in a report Thursday.

Twitter removed 45.4% of the hateful posts it was notified about in a sample this year, up from 49.8% in 2021, European Union officials. wrote in his report.

Twitter performed worse on that metric than any other social media platform tested, according to the report, but some of them, including Facebook, Instagram and TikTok, also lagged behind a year earlier.

YouTube improved, removing 90.4% of reported posts, up from 58.8% a year earlier, according to the report.

The data was collected from March to May, months before tech mogul Elon Musk bought Twitter for $44 billion and began to further relax the site’s enforcement of hateful posts.

Musk announced Thursday that he would loosen the enforcement further, tweeting that he would grant blanket «amnesty» next week to accounts that Twitter had previously suspended.

Musk’s policies have put Twitter on a potential collision course with the EU, where hate speech does not have the protection from government action that it does in the United States under the First Amendment. A new EU law, the Digital Services Act, threatens tech companies with billions of dollars in fines if they don’t strictly police their platforms.

Didier Reynders, the EU justice commissioner, said the latest data could be used to apply the new law.

“Last year, I called on companies to reverse the general downward trend in notification and action without delay. This has yet to fully happen – companies clearly need to step up their engagement,” Reynders said in a statement.

Twitter, YouTube and Meta, the parent company of Facebook and Instagram, did not immediately respond to requests for comment on the report, which EU officials released on Thanksgiving Day, when most US offices. USA are closed.

TikTok said in a statement that the EU research is «valuable for sharing knowledge and finding new ways we can improve our policies and strengthen our enforcement.» We look forward to continuing our cooperation with the European Commission, NGOs and other signatories to maintain TikTok. a safe, positive and inclusive place for creative expression as we tackle the complex and ever-evolving issue of hate speech.»

Racist tweets spiked rapidly after Musk completed his purchase of Twitter in late October, outside researchers said.

Musk has said that he has focused less on removing hateful posts and more on limiting how often people see those posts, preventing them from going viral. In a tweet Wednesday, Musk said such opinions, or «impressions,» of hate speech dropped by a third before he bought the company; external researchers have not verified that claim.

Twitter’s rulebook has long banned posts promoting «hateful conduct,» and that policy was still on their website Thursday.

But Musk has also fired or fired a large part of Twitter’s workforce in his four weeks as owner and CEO, and the cuts have included people whose job it was to watch for content that violated Twitter rules.

CNBC reported on Wednesday that tensions are also brewing between Twitter and the companies that run the two largest app stores, Apple and Google, which have their own rules on content moderation.

EU officials said they worked with 33 civil society organizations and three public bodies to notify tech companies of breaches and monitor takedowns.

It is the seventh annual report of its kind that they have published since 2016.