Meta’s 2025 policy changes linked to rise in antisemitic content under scrutiny
Consensus Summary
Meta’s January 2025 policy changes to reduce 'censorship' by shifting from proactive to reactive enforcement of hate speech have faced scrutiny during the Royal Commission on Antisemitism and Social Cohesion. Both sources agree that Meta’s global director Benjamin Good testified on Monday, denying the changes led to more antisemitic content, despite a 79% drop in hate speech removals from platforms like Facebook and Instagram. Data shows Facebook actions on hate speech fell from 5.8 million items in late 2024 to 1.2 million in mid-2025, and Instagram actions dropped from 7.4 million to 2 million in the same periods. Meta’s internal documents reveal a nuanced approach, allowing offensive but non-harmful speech while prohibiting attacks on protected characteristics, including coded antisemitic language like 'Zionists control the media'. The Guardian highlights Meta’s internal FAQs explicitly stating it does not police offensiveness, while the ABC focuses on Meta’s commitment to learning from missed antisemitic posts during the hearings. Both sources agree the changes were intended to balance free speech with over-enforcement risks, though critics argue the shift has enabled more hate speech to spread unchecked.
✓ Verified by 2+ sources
Key details reported by multiple sources:
- Meta’s global director of core policy Benjamin Good testified before the Royal Commission on Antisemitism and Social Cohesion on Monday
- Meta announced sweeping changes to its Community Standards on Hate Speech in January 2025, shifting to a reactive enforcement model relying on user-reported violations
- Meta’s platforms (Facebook, Instagram, Threads) have 3.5 billion users combined
- Facebook removed 5.8 million items for hate speech from October 2024 to December 2024, dropping to 1.2 million from July 2025 to September 2025
- Instagram removed 7.4 million items for hate speech from October 2024 to December 2024, dropping to 2 million from July 2025 to September 2025
- Meta’s hateful conduct policy violations metric dropped by 79% after January 2025 changes
- Meta’s internal documents state offensive comments like 'immigrants are criminals' or 'white people are all Nazis' are allowed under new policies
- Meta prohibits attacks based on protected characteristics, including dehumanizing speech, slurs, and harmful stereotypes
- Meta expanded its hate speech policies to include prohibiting claims like 'Zionists control the media' as antisemitic coded language
- Meta’s hateful conduct policy violations metric has remained at 0.02% since 2022
Points of Difference
Details reported by only one source:
- Meta’s Mark Zuckerberg announced changes in January 2025 to reduce over-enforcement, stating 'we're going to catch less bad stuff, but we'll also reduce the number of innocent people's posts and accounts that we accidentally take down'
- Four antisemitic posts on Facebook during the royal commission hearings were initially not removed but later deleted after review
- Meta’s proactive AI tools were primarily used for high-severity violations like terrorism, drugs, fraud, and scams
- Meta’s core policy director Benjamin Good personally reviewed the four missed antisemitic posts and committed to learning from them
- Meta’s internal documents suggest responses to questions about policy changes, emphasizing a distinction between offensive speech and content leading to offline harm
- Meta’s policy changes were not aimed at altering its approach to antisemitism but to allow political debates on topics like gender and immigration
- Meta’s policy changes were announced after Donald Trump’s re-election in 2025, described as a 'trade-off' to reduce 'censorship'
- Meta’s internal FAQs state 'it is not Meta’s role to police offensiveness' and allow phrases like 'gay people are sinners' or 'black people are more violent than whites'
- Meta’s Facebook Australia public policy director Mia Garlick confirmed Australia retains factcheckers for misinformation, unlike other regions
- Meta’s internal documents note that 'black people are all drug dealers' is prohibited as it involves specific criminal behavior, unlike 'black people are more violent than whites'
- Meta’s counsel assisting Richard Lancaster described the policy changes as 'entirely unrealistic' to not impact moderation of antisemitic content
Contradictions
Conflicting information between sources:
- The ABC states Meta’s policy changes were announced in January 2025, while the Guardian implies the changes were directly tied to Donald Trump’s re-election in 2025, though both sources agree on the January 2025 timeline
- The ABC mentions Meta’s proactive AI tools were used for high-severity violations, while the Guardian does not reference this specific detail
- The ABC notes Meta’s internal documents suggest responses to questions about policy changes, while the Guardian focuses on the internal FAQs explicitly stating Meta does not police offensiveness
- The Guardian emphasizes Meta’s internal FAQs allow phrases like 'white people are all Nazis' as acceptable, while the ABC does not explicitly mention this example in its internal documents
Source Articles
Meta exec rejects moderation changes led to more antisemitism
A Meta executive has rejected claims a change in the company's policies led to a rise in antisemitism while giving evidence to the royal commission.
Meta bosses grilled over decision to cut ‘censorship’ that has potentially unleashed more antisemitic content
‘Over-enforcement poses significant risk to the communities that we try to protect,’ royal commission told A decision aimed at reducing “censorship” on major social media sites including Facebook and Instagram potentially led to greater levels of hate speech, the royal commission into antisemitism has heard. Meta announced in January 2025 , after the re-election of Donald Trump in the US, that it would “reduce censorship”, get rid of factcheckers and only tackle illegal and very serious violatio