Meta's Policy Director Defends Changes Amidst Antisemitic Content Concerns
In a recent hearing, Meta's policy director, Benjamin Good, faced scrutiny over the company's controversial policy changes and their impact on content moderation. The focus was on the potential increase in antisemitic content following the adjustments to Meta's Community Standards on Hate Speech.
Good defended the changes, arguing that they aimed to strike a balance between proactive and reactive content moderation. He explained that the previous 'gold standard' approach, which relied heavily on artificial intelligence to proactively remove harmful posts, had led to over-enforcement and the accidental removal of innocent accounts. The new policy, he claimed, prioritizes a more reactive approach, where users are encouraged to report potentially harmful content.
However, the effectiveness of this shift is questionable. Data reveals a significant drop in hate speech content removal post-policy change. From October 2024 to December 2024, Meta removed approximately 5.8 million items of hate speech, but this number plummeted to 1.2 million from July to September 2025. Similarly, Instagram's hate speech removals decreased from 7.4 million to 2 million during the same period.
The Royal Commissioner, Virginia Bell, pressed Good on these figures, questioning whether the policy change was the sole reason for the decline. Good acknowledged the drop but remained cautious about attributing it solely to the policy shift. He suggested that the reactive approach still holds value, as reported content is more likely to be removed, even if it's not initially flagged by Meta's AI.
The hearing also highlighted instances where antisemitic posts slipped through the cracks. The royal commission brought to light at least four such posts that were initially deemed not to violate community standards but were later removed after a review. Good acknowledged the oversight and vowed to learn from these instances, emphasizing the importance of monitoring potential harm, even if it means missing some content.
In conclusion, while Good defended the policy changes, the hearing revealed ongoing challenges in content moderation. The company's approach to addressing hate speech, particularly antisemitism, remains under scrutiny, and the balance between proactive and reactive measures is a delicate one that Meta continues to navigate.
