Meta has systematically monitored pro-Palestinian content around the world since October 7, according to a new 51-page report from Human Rights Watch.
The group has documented more than 1,000 cases of Facebook and Instagram content being removed and accounts suspended or permanently suspended.
Meta identified six main patterns of unjustified censorship of content in support of Palestine and Palestinians, each of which occurred in at least 100 cases.
Forms of censorship include removal of posts, stories, and comments, suspension or deletion of accounts, inability to interact with content, inability to follow or mention accounts, limited use of live streaming features, and Ghost bans.
Human Rights Watch cited content written by users from more than 60 countries, most of it in English and all related to supporting peace in Palestine, and expressed in different ways.
Human Rights Watch said Meta classified the group's messages as spam because it was looking for examples of censorship.
The organization said in the report: “The censorship of Palestine-related content on Facebook and Instagram is systematic and global, and Meta’s inconsistent application of its policies has led to the wrongful removal of Palestine-related content.”
The group said the problem is due to incorrect and inconsistent implementation of definitional policies, over-reliance on automated content moderation tools, and undue government influence over content removal procedures.
In a statement to The Guardian, Mehta admitted that she had made frustrating mistakes, but explained that deliberately and systematically silencing the voice was wrong.
“The idea that a thousand examples from the rich content of these incidents is evidence of systematic analysis might be a good title for the report, even if it is misleading,” Mehta said.
Meta confirmed that it is the only company in the world that has publicly declared its commitment to respecting human rights on issues related to Israel and Palestine.
The company added: “This report ignores the reality of implementing our policies around the world during highly polarized and rapidly evolving conflicts, which contributed to increasing the report’s content. Our policies are designed to be accessible to everyone. Have a voice while ensuring the security of our platform.
It is worth noting that this is the second time this month that Meta has faced accusations of regularly removing pro-Palestinian content and voices.
Last week, Senator Elizabeth Warren wrote to Meta co-founder and CEO Mark Zuckerberg after hundreds of Instagram users reported since October that content had been removed and requested information after their accounts were suspended.
Meta's board said the company made a mistake by removing the videos from Instagram and Facebook, saying the videos were useful in informing the world of human suffering on both sides.
Meta users documented what they described as technological bias against pro-Israel content and against pro-Palestinian posts.
The Instagram translator replaced the English word “Palestinian,” followed by the Arabic phrase “Praise God,” with the English word “Palestinian Terrorist.”
When WhatsApp's AI was asked to generate images of Palestinian girls and boys, it generated images of children carrying weapons, while images of Israeli children did not contain weapons.