Meta’s Oversight Board has considered its first Threads case and overturned the company’s initial decision and first appeal. In regards to a post about outgoing Japanese Prime Minister Fumio Kishida that used a phrase translated into English as “die/go die,” the board determined that the phrase was used figuratively and not as a literal threat or call for violence.
The case was initiated by a Threads post that featured a news article about Kishida and his response to his political party’s (ahem) “fundraising irregularities.” The caption criticized the prime minister, accusing him of tax evasion.
The user’s reply sought an explanation from the government leader and used the phrase “死ね,” or “die/go die,” calling him a tax evader. The post also included the phrase “haha” and derogatory language about people who wear glasses. (Watch yourself there, partner!)
The post went largely ignored, receiving no likes. But someone reported it under Meta’s bullying and harassment rules. After three weeks, a reviewer at Meta found that it violated rules on violence and incitement.
The user appealed, and a second reviewer agreed with the first that it violated policy. Another appeal took the issue to the board, which accepted the case and overruled the two human reviewers who had removed it.
“In this case, the threat against a political leader was intended to draw attention to alleged corruption in the form of non-literal political criticism using harsh language, which is not uncommon on Japanese social media,” Meta’s oversight board wrote in its explanation. “It was unlikely to cause harm.” The board considered the poster’s use of “haha” to help determine its figurative meaning.
The board said the moderators who removed the post, despite speaking Japanese and understanding local content, were making a “mistake.” It recommends Meta clarify its internal guidelines and provide reviewers “more guidance on how to evaluate language and local content.”
Meta’s oversight board said the violence and incitement policy, which includes a rule prohibiting the phrase “death” against “high-risk individuals,” is not clear enough. It said that while the company’s policy reasoning suggests that context matters in threat assessment, its reviewers do not have the authority to assess cases involving the phrase “death.”
The board reiterated its 2022 recommendation for Meta to clarify that rhetorical threats using the phrase are “generally permitted, except when directed at high-risk individuals, and to provide criteria for when threatening statements directed at heads of state are permitted to protect rhetorical political speech.”
In addition, the board recommended that Meta clarify how the policy differs for “public figures” versus “high-risk individuals.” This clears up confusion that threats against public figures are only removed if they are “credible.”
In contrast, threats against others are removed “regardless of credibility.” The Oversight Board had a busy September after deciding on only 53 cases last year. Last week, it ruled that the phrase “from the river to the sea” should not be banned and, in a case bearing some resemblance to this one, it distinguished death threats in Venezuela from “aspirational statements.”