In a leaked Commission report, firms deleted 62.5 percent of material after it was flagged, down from 71 percent in the previous annual audit.

The world's largest social media companies removed less hate speech from their platforms in 2021 compared to last year, according to the European Commission's annual review of the firms' content moderation activities, seen by POLITICO.

The yearly checkup on how Facebook, Google and others handle everything from misogynistic online posts to digital abuse targeting the LGBTQ+ community found that social media companies deleted 62.5 percent of such flagged material, over a 6-week period between March 1 and April 2021. That compares to a 71 percent removal rate when Brussels conducted its last review in late 2019.

The European Union is in the middle of a political fight on overhauling how it approaches online content regulation, with the European Parliament working on its revised version of the Digital Services Act. Those proposals would impose hefty fines on companies that don't combat illegal material, including hate speech, as well as require greater transparency over how specific posts were displayed in people's social media feeds.

The battle intensified this week after Frances Haugen, a former Facebook employee, told U.S. lawmakers the tech giant put its own business interests over the need to keep its 2.4 billion users safe online. Mark Zuckerberg, the company's chief executive, rejected those claims. Commissioners Věra Jourová and Thierry Breton both met virtually with Haugen on Wednesday, with the Czech politician telling a POLITICO event last week that Brussels should work with Washington on tackling these challenges.

"From the point of view of spreading misinformation, foreign interference, cyberattacks, we have the same adversaries," she said.

As part of the EU's so-called code of conduct on countering illegal hate speech online — voluntary standards dating from 2016 — social media companies are audited annually to determine how they are reducing the spread of this material across their platforms. The initial valuation, from late 2016, found that Google's YouTube, Facebook and Twitter all removed less than half of the material that was flagged to them as problematic.

Those figures have steadily increased over the years as the companies responded to regulatory pressure to act. Companies that participated in this year's audit included Facebook, YouTube, Twitter, Instagram and TikTok. Critics, though, say these firms are still not addressing the widespread existence of harmful material on their platforms, particularly in non-English languages, and that hate speech can still spread easily in vast parts of these networks.

As part of the most recent review, which will be published later on Thursday, 81 percent of content submitted by outside organizations was reviewed within the first 24 hours. That compares to a 90 percent review rate during the previous audit. In total, 35 groups across 22 EU countries participated in the audit, which included just over 4,500 notifications of potential hate speech being sent, collectively, to all of the platforms.

Both Facebook and YouTube reported a decline in how much content they had removed over the 6-week auditing period. Facebook deleted 70 percent of the material, compared to 88 percent in the previous review. YouTube's figures dropped almost 21 percentage points, to 59 percent of material that was removed, according to the analysis obtained by POLITICO. Both companies have increasingly turned to automated systems to police online content, particularly during the COVID-19 pandemic, which has led to more content in some instances being removed, but fewer material in other areas being deleted.

In contrast, Twitter and Facebook's Instagram both reported increases in the amount of hate speech they had removed once notified, according to the Commission's audit. TikTok — the Chinese-owned app that was participating in its first audit — said that it had removed 80 percent of such material, the highest figure of any of the major platforms that participated in this year's voluntary review.

“We don’t tolerate hate speech on Facebook or Instagram," a Facebook spokesperson said in a statement, adding that the company had yet to read the Commission's report. "We’re finding and removing more hate speech on our platforms than ever before, and reducing the prevalence of hate speech."

BY MARK SCOTT

Social media companies remove less hate speech in 2021 – POLITICO