Facebook won't say if its algorithms boosted Trump's violent rhetoric

[May 06, 2021]  By Helen Coster and Paresh Dave

(Reuters) - Facebook Inc's oversight board said on Wednesday that the company did not answer questions about whether its algorithms amplified inflammatory posts by then-U.S. President Donald Trump and contributed to the deadly siege on the Capitol in January.

The board recommended that Facebook review how it might have potentially contributed to the violence and the false narrative of election fraud.

Many Democrats and other critics have said Trump’s posts helped fuel the attack, which led to five deaths. Facebook indefinitely banned Trump from posting in the wake of the violence and asked for further guidance from the oversight board, a 20-person panel funded by the company to review content moderation decisions.

The board found Facebook’s indefinite suspension of Trump’s account was arbitrary because it did not follow a clear published procedure. It called on the company to develop new rules within six months that would lead to either Trump’s reinstatement or some other penalty. Kicking the decision on what to do with Trump back to Facebook drew condemnation from both sides of the political spectrum.

The company’s role in promoting posts by Trump is important to understand, the board said, because measures short of banning his account could have been enough to limit the risk of violence. Facebook for years has been criticized for designing its News Feed algorithms in ways that promote divisive and inflammatory content.

But the question about whether any internal analysis had been conducted since January was among seven out of 46 questions that Facebook declined to answer.

A 3D-printed Facebook logo is seen placed on a keyboard in this illustration taken March 25, 2020. REUTERS/Dado Ruvic/Illustration

"This makes it difficult for the Board to assess whether less severe measures, taken earlier, may have been sufficient to protect the rights of others," the decision stated.

The board did not describe possible alternatives to suspending Trump, but Facebook and other social media companies have previously blocked users from re-sharing some problematic posts on their services or shown warnings to viewers before or next to such content.

Yael Eisenstat, a researcher on technology and democracy who was previously global head of elections integrity operations for political ads at Facebook, said it was clear the social network's features played a role in sparking the attack.

"We're diverting time and attention away from holding Facebook accountable for its designs and policies," Eisenstat told reporters on Wednesday of the focus on Trump's ban.

Suspending heads of state or high-ranking government officials that have repeatedly posted problematic messages is fine because they “have a greater power to cause harm than other people,” according to the board. But the suspension should be “for a determinate period sufficient to protect against imminent harm,” which could include a permanent ban, the board concluded.

BuzzFeed reported last month that Facebook recently created a report on how it failed to curb "Stop the Steal" groups on Facebook that falsely claimed widespread election fraud. Facebook told BuzzFeed it was not a definitive report.

(Reporting by Helen Coster and Paresh Dave; Additional reporting by Sheila Dang; Editing by Lisa Shumaker)

[© 2021 Thomson Reuters. All rights reserved.]

Copyright 2021 Reuters. All rights reserved. This material may not be published, broadcast, rewritten or redistributed.  Thompson Reuters is solely responsible for this content.