Meta’s review of hate speech policy sparks concern of further censorship of pro-Palestinian content | Meta

Meta is considering expanding and “revisiting” its hate speech policy around the term “Zionist”, the Guardian has confirmed. The company reached out to and met with more than 10 Arab, Muslim and pro-Palestinian organizations on Friday to discuss the company’s plans to review the policy to ensure the term “Zionist” is not being used as a proxy for Jewish or Israeli people, according to an email the Guardian reviewed.

The policy as it exists allows “Zionist” to be used in “political discourse but removed when it’s used explicitly as a proxy for Jews or Israelis in a dehumanizing or violent way”, according to an email a Meta representative sent to the organizations inviting them to the Friday meeting. The email further stated the company was considering reviewing it in light of posts users and “stakeholders” have recently reported, the Meta representative wrote. The Intercept first reported the potential change in policy.

In an email to another organization, the Meta representative said the company’s current policies don’t allow users to attack others based on their protected characteristics such as nationality or religion and that the policy “requires a current understanding of how people use language to reference those characteristics”.

“While the term ‘Zionist’ often refers to a person’s ideology, which is not a protected characteristic, it can also be used to refer to Jewish or Israeli people,” the email read. “Given the increase in polarized public discourse due to current events in the Middle East, we believe it’s important to assess our guidance for reviewing posts that use the term ‘Zionist.’”

Organizations involved in discussions, which include MPower Change, 7amleh and Jewish Voice for Peace, expressed deep concern in the meeting about whether this change would further censor pro-Palestinian voices. Several reports commissioned by 7amleh and Human Rights Watch as well as one commissioned by Meta confirmed that Palestinian accounts have long been systematically silenced and stifled on Meta-owned platforms.

However, Meta asserts that an independent consultancy it works with, BSR, “did raise important concerns around under-enforcement of content … and specific instances where they considered our policies and processes had an unintentional impact on Palestinian and Arab communities”.

Linda Sarsour, the executive director of Muslim advocacy organization MPower Change, said Meta’s director of content policy stakeholder engagement, Peter Stern, provided few details about why the company was revisiting the policy now and how it would be implemented or enforced in a way that doesn’t stifle political expression. “If you already have a policy that’s addressing Zionism as a proxy, then why are we having this conversation? Why is there further consideration to expand this policy?” Sarsour said.

The groups also questioned how such policies would be enforced, including whether humans or Meta’s algorithms – which have been criticized for unfairly censoring Palestine-related content – would be employed to detect and censor such language.

“The AI-powered systems to initially flag posts are problematic – there is no human review until it’s too late,” said Abed Ayoub, the national executive director of the American-Arab Anti-Discrimination Committee. The organization was one of those in attendance at the meeting with Meta.

Before the meeting, 73 organizations sent a letter to Meta saying the proposed expansion of the policy would “too easily mischaracterize conversations about Zionists – and by extension, Zionism – as inherently antisemitic … treating ‘zionist’ as a proxy will also encourage the incorrect and harmful conflation of criticism of the acts of the state of Israel with antisemitism.”

“This move will prohibit Palestinians from sharing their daily experiences and histories with the world, be it a photo of the keys to their grandparent’s house lost when attacked by Zionist militias in 1948, or documentation and evidence of genocidal acts in Gaza over the past few months, authorized by the Israeli Cabinet, which includes members of the Religious Zionist Party,” the letter read. “And it would prevent Jewish users from discussing their relationships to Zionist political ideology.”

Sarsour said Meta presented examples of posts that would be taken down. They included a post that called Zionists rats – “that would obviously be taken down because of the historical trope and demonization of Jewish people as rats”, Sarsour said. An example of a post about Zionists taking someone’s land would not get taken down, she recalled from the presentation.

These specific content-moderation decisions would be agreeable to Sarsour’s organization, but Meta so far has not had a reliable track record of protecting Palestinian and pro-Palestinian users and speech, Sarsour said.

In their letter, organizations also expressed concern about a lack of response to the surge in censorship of pro-Palestinian content as well as hate speech, which they say is at an all-time high since the war in Gaza began. There have been extensive reports of Meta suppressing pro-Palestinian content since the beginning of the most recent conflict, as well as in the past. An independent analysis commissioned by Meta concluded in 2022 that the company’s social networks had violated Palestinian human rights by censoring content related to Israel’s attacks on Gaza. In December 2023, the senator Elizabeth Warren demanded that Mark Zuckerberg answer questions regarding reports that such content was repeatedly being censored or suppressed following the Hamas attacks in October 2023 that escalated the recent conflict.

There’s no effort parallel to the proposal on Zionism-related language to protect Palestinians, the organizations reiterated, and expanding the censorship of posts about Zionists and Zionism will not effectively dismantle antisemitism.

“Not only will this proposal be ineffective at combating antisemitism, it entirely ignores the digitally fueled oppression of Palestinians at a time when multiple courts and human rights experts around the world have affirmed the plausibility that a genocide is happening in Gaza and ordered all international actors to do whatever they can to prevent further harm,” the letter read.

In response to a request for comment, Corey Chambliss, a spokesperson from Meta, shared the earlier statement regarding the “increase in polarized public discourse”. He added that Meta is investigating whether and how it could scale a nuanced response to such language and that the company will have ongoing discussions with stakeholders to refine its policy.

The policy discussion comes at a high-stakes time in an ongoing conflict, during which accurate information and its dissemination can have widespread repercussions. More than 25,000 Palestinians have been killed since the attack on Gaza began in October 2023.

“To implement a policy like this in the middle of a genocide is very problematic,” said Ayoub, who added that his organization has fielded reports of censorship and suspended accounts from a number of users of Meta platforms. “Adding this policy on top of that is going to make things 10 times worse – it’s going to cause untold damage to our community.”

Meta is considering expanding and “revisiting” its hate speech policy around the term “Zionist”, the Guardian has confirmed. The company reached out to and met with more than 10 Arab, Muslim and pro-Palestinian organizations on Friday to discuss the company’s plans to review the policy to ensure the term “Zionist” is not being used as a proxy for Jewish or Israeli people, according to an email the Guardian reviewed.

The policy as it exists allows “Zionist” to be used in “political discourse but removed when it’s used explicitly as a proxy for Jews or Israelis in a dehumanizing or violent way”, according to an email a Meta representative sent to the organizations inviting them to the Friday meeting. The email further stated the company was considering reviewing it in light of posts users and “stakeholders” have recently reported, the Meta representative wrote. The Intercept first reported the potential change in policy.

In an email to another organization, the Meta representative said the company’s current policies don’t allow users to attack others based on their protected characteristics such as nationality or religion and that the policy “requires a current understanding of how people use language to reference those characteristics”.

“While the term ‘Zionist’ often refers to a person’s ideology, which is not a protected characteristic, it can also be used to refer to Jewish or Israeli people,” the email read. “Given the increase in polarized public discourse due to current events in the Middle East, we believe it’s important to assess our guidance for reviewing posts that use the term ‘Zionist.’”

Organizations involved in discussions, which include MPower Change, 7amleh and Jewish Voice for Peace, expressed deep concern in the meeting about whether this change would further censor pro-Palestinian voices. Several reports commissioned by 7amleh and Human Rights Watch as well as one commissioned by Meta confirmed that Palestinian accounts have long been systematically silenced and stifled on Meta-owned platforms.

However, Meta asserts that an independent consultancy it works with, BSR, “did raise important concerns around under-enforcement of content … and specific instances where they considered our policies and processes had an unintentional impact on Palestinian and Arab communities”.

Linda Sarsour, the executive director of Muslim advocacy organization MPower Change, said Meta’s director of content policy stakeholder engagement, Peter Stern, provided few details about why the company was revisiting the policy now and how it would be implemented or enforced in a way that doesn’t stifle political expression. “If you already have a policy that’s addressing Zionism as a proxy, then why are we having this conversation? Why is there further consideration to expand this policy?” Sarsour said.

The groups also questioned how such policies would be enforced, including whether humans or Meta’s algorithms – which have been criticized for unfairly censoring Palestine-related content – would be employed to detect and censor such language.

“The AI-powered systems to initially flag posts are problematic – there is no human review until it’s too late,” said Abed Ayoub, the national executive director of the American-Arab Anti-Discrimination Committee. The organization was one of those in attendance at the meeting with Meta.

Before the meeting, 73 organizations sent a letter to Meta saying the proposed expansion of the policy would “too easily mischaracterize conversations about Zionists – and by extension, Zionism – as inherently antisemitic … treating ‘zionist’ as a proxy will also encourage the incorrect and harmful conflation of criticism of the acts of the state of Israel with antisemitism.”

“This move will prohibit Palestinians from sharing their daily experiences and histories with the world, be it a photo of the keys to their grandparent’s house lost when attacked by Zionist militias in 1948, or documentation and evidence of genocidal acts in Gaza over the past few months, authorized by the Israeli Cabinet, which includes members of the Religious Zionist Party,” the letter read. “And it would prevent Jewish users from discussing their relationships to Zionist political ideology.”

Sarsour said Meta presented examples of posts that would be taken down. They included a post that called Zionists rats – “that would obviously be taken down because of the historical trope and demonization of Jewish people as rats”, Sarsour said. An example of a post about Zionists taking someone’s land would not get taken down, she recalled from the presentation.

These specific content-moderation decisions would be agreeable to Sarsour’s organization, but Meta so far has not had a reliable track record of protecting Palestinian and pro-Palestinian users and speech, Sarsour said.

In their letter, organizations also expressed concern about a lack of response to the surge in censorship of pro-Palestinian content as well as hate speech, which they say is at an all-time high since the war in Gaza began. There have been extensive reports of Meta suppressing pro-Palestinian content since the beginning of the most recent conflict, as well as in the past. An independent analysis commissioned by Meta concluded in 2022 that the company’s social networks had violated Palestinian human rights by censoring content related to Israel’s attacks on Gaza. In December 2023, the senator Elizabeth Warren demanded that Mark Zuckerberg answer questions regarding reports that such content was repeatedly being censored or suppressed following the Hamas attacks in October 2023 that escalated the recent conflict.

There’s no effort parallel to the proposal on Zionism-related language to protect Palestinians, the organizations reiterated, and expanding the censorship of posts about Zionists and Zionism will not effectively dismantle antisemitism.

“Not only will this proposal be ineffective at combating antisemitism, it entirely ignores the digitally fueled oppression of Palestinians at a time when multiple courts and human rights experts around the world have affirmed the plausibility that a genocide is happening in Gaza and ordered all international actors to do whatever they can to prevent further harm,” the letter read.

In response to a request for comment, Corey Chambliss, a spokesperson from Meta, shared the earlier statement regarding the “increase in polarized public discourse”. He added that Meta is investigating whether and how it could scale a nuanced response to such language and that the company will have ongoing discussions with stakeholders to refine its policy.

The policy discussion comes at a high-stakes time in an ongoing conflict, during which accurate information and its dissemination can have widespread repercussions. More than 25,000 Palestinians have been killed since the attack on Gaza began in October 2023.

“To implement a policy like this in the middle of a genocide is very problematic,” said Ayoub, who added that his organization has fielded reports of censorship and suspended accounts from a number of users of Meta platforms. “Adding this policy on top of that is going to make things 10 times worse – it’s going to cause untold damage to our community.”