AI Analysis Reveals Doubling of Antisemitic Tweets on Twitter After Musk Takeover

Thumbnail Image

The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.

Researchers using advanced AI and machine learning models found that antisemitic tweets on Twitter more than doubled—rising by 105%—after Elon Musk acquired the platform. The AI systems detected a sustained increase in hate speech, highlighting the impact of platform management and moderation policies on harmful content proliferation.[AI generated]

Why's our monitor labelling this an incident or hazard?

The article explicitly mentions the use of a machine learning-based hate speech detection tool (Beam) to identify antisemitic tweets, confirming AI system involvement. The harm described is the significant increase in antisemitic content and hate speech on the platform, which harms communities and contributes to offline violence. Although the AI system is used for detection rather than causing the harm, the event centers on the consequences of AI system use in content moderation and the resulting harm. The rise in hate speech and its societal impact meets the criteria for harm to communities (d). Therefore, this qualifies as an AI Incident due to the direct link between AI system use in content detection and the documented harm.[AI generated]
Industries
Media, social platforms, and marketing

Harm types
Other

Business function:
Research and development

AI system task:
Event/anomaly detection


Articles about this incident or hazard