Facebook's automated spam filter system malfunctioned and blocked users from posting legitimate news articles about coronavirus from multiple reputable outlets including Business Insider, BuzzFeed, The Atlantic, and The Times of Israel.
On Tuesday, March 17, 2020, Facebook users reported being unable to post articles from several news outlets including Business Insider, BuzzFeed, The Atlantic, and The Times of Israel. The posts were being flagged as spam by Facebook's automated content moderation system. This occurred during the COVID-19 pandemic when Facebook had sent many of its human content moderators home and was relying more heavily on automated software for content moderation. Former Facebook security executive Alex Stamos speculated that the reduced human oversight might be contributing to the malfunction. Facebook's vice president of integrity, Guy Rosen, acknowledged the issue as a bug in an anti-spam system and denied it was related to changes in the content moderator workforce. The company stated they were working to fix the problem and restore the blocked posts. Multiple users complained on Twitter that their posts about coronavirus news were being incorrectly flagged as spam, with some noting that articles critical of Trump were among those removed.
Domain classification, causal taxonomy, severity scores, and national security assessments were LLM-classified and may contain errors.
AI systems that fail to perform reliably or effectively under varying conditions, exposing them to errors and failures that can have significant consequences, especially in critical applications or areas that require moral reasoning.
AI system
Due to a decision or action made by an AI system
Unintentional
Due to an unexpected outcome from pursuing a goal
Post-deployment
Occurring after the AI model has been trained and deployed