The information displayed in the AIM should not be reported as representing the official views of the OECD or of its member countries.
Amnesty International Brazil tested Facebook's AI moderation by submitting 15 paid ads, 14 containing electoral disinformation. The AI system failed to block over half, accepting them for future publication. Although the ads were not published, the failure exposes a significant risk of disinformation spreading during Brazil's elections.[AI generated]
Why's our monitor labelling this an incident or hazard?
The article explicitly involves an AI system used by Meta for moderating ads on Facebook. The system's failure to detect disinformation and harmful political content directly contributes to misinformation spreading, which harms communities and undermines democratic processes. This fits the definition of an AI Incident because the AI system's malfunction has directly led to harm (misinformation and potential electoral harm).[AI generated]