Facebook’s Oversight Board has issued its first round of rulings, upholding one removal and overturning four decisions involving hate speech, nudity, and misinformation. Together, the rulings take an expansive view of what users can post under the current policies, based on concerns about vague rules and protecting freedom of expression online.
The Oversight Board — composed of experts outside Facebook — accepted its first set of cases in December. While the original slate included six incidents, a user in one case proactively deleted their post, rendering the decision moot. Facebook has pledged to follow its rulings within seven days and respond to recommendations for new policies within 30 days. In a response to the rulings, Facebook said it had already restored all the content in question.
One case objected to Facebook’s automated takedown system
The five cases covered posts across four continents. In Brazil, the board ruled in favor of a woman whose Instagram post about breast cancer was automatically removed for nudity. Facebook had already restored the photo, but the board objected to the initial removal, saying the fully automated decision “indicates the lack of proper human oversight which raises human rights concerns.”
Two other cases show the bounds of what the board considers hate speech. A panel upheld Facebook removing a Russian post with a demeaning slur against Azerbaijani people. But it overturned a decision in Myanmar, saying that while the post “might be considered offensive, it did not reach the level of hate speech.”
The post was written in Burmese, and the decision was based on some fine translation differences. Facebook initially interpreted it as saying “[there is] something wrong with Muslims psychologically,” but a later translation rendered it as “[specific] male Muslims have something wrong in their mindset,” which was deemed “a commentary on the apparent inconsistency between Muslims’ reactions to events in France and in China.”
A post about Muslims in Myanmar was considered “offensive” but “did not reach the level of hate speech”
As the Facebook board acknowledges, Myanmar is in the grips of an ongoing genocide against the Rohingya Muslim minority, incited partly through inflammatory Facebook posts. However, it declared that “statements referring to Muslims as mentally unwell or psychologically unstable are not a strong part of this rhetoric,” and “while the post might be considered pejorative or offensive towards Muslims, it did not advocate hatred or intentionally incite any form of imminent harm.”
Other decisions hinge on Facebook explaining its policies badly, rather than the specific content of the post. A US-based post, for instance, compared a quote from Nazi propaganda leader Joseph Goebbels to American political rhetoric. Facebook determined it violated hate speech policies because it didn’t explicitly condemn Goebbels, but “Facebook is not sufficiently clear that, when posting a quote attributed to a dangerous individual, the user must make clear that they are not praising or supporting them,” the board said.
The board still hasn’t decided whether Facebook could ban Trump
Another case, from France, referred falsely to hydroxychloroquine as a “cure” for COVID-19. But the reference was part of a comment about government policies, not an encouragement to take the drug, and the board said this didn’t rise to the level of causing “imminent harm.” The board said that Facebook’s rules about medical misinformation were “inappropriately vague and inconsistent with international human rights standards,” and it’s encouraged Facebook to publish clearer guidelines about what counts as “misinformation,” as well as a transparency report about how it has moderated COVID-19-related content.
Facebook suggests it won’t take a softer line on COVID-19 misinformation
Facebook says it will apply the precedent from these rulings to similar content on the network, although it didn’t give a specific number of posts that were affected. It’s still formulating policy changes, but it discussed the medical misinformation case specifically, saying that its takedown approach “will not change” while the pandemic is ongoing. However, it plans to publish updated COVID-19-related policies soon. “It is critical for everyone to have access to accurate information, and our current approach in removing misinformation is based on extensive consultation with leading scientists, including from the CDC and WHO,” writes content policy vice president Monika Bickert.
The Oversight Board says it will soon take on a new slate of cases, which can be drawn from user appeals or referred directly by Facebook. It will also open a public comment period for its highest-profile case so far: whether Facebook and Instagram should indefinitely suspend former President Donald Trump.
Facebook’s Oversight Board — effectively a “supreme court” for the social network — was criticized for a slow rollout after its initial announcement last year. A separate group of activists calling themselves the “Real Facebook Oversight Board” have also called it too narrowly focused on putting content back online, rather than addressing whether Facebook should moderate more strictly.
Stanford Cyber Policy Center co-director Nate Persily noted that individual decisions aren’t the only thing at stake in this set of rulings. “The results in these decisions are less important than the signals/precedent set for how the board will operate, how it considers its jurisdiction, what info about [Facebook] and its posts will be revealed in the decisions, and how ambitious the Board will be in checking Facebook,” he tweeted after the ruling.
Similar to a national Supreme Court, the Oversight Board’s decisions are supposed to help clarify Facebook’s complicated rules. Unlike a democratic nation, however, the company can easily change its own moderation policies, and Facebook is under no legal obligation to abide by the board’s rulings.