Meta Shifts to AI for Risk Assessments, Raising Concerns

Meta plans to automate up to 90% of risk assessments for new features and updates on its platforms like Facebook and Instagram, replacing human evaluators with AI systems that will determine potential privacy violations, harm to minors, and spread of misinformation, raising concerns about oversight and unintended consequences.

• Meta (formerly Facebook) plans to automate up to 90% of risk assessments for new product features and updates using AI systems instead of human evaluators.
• Current and former employees express concerns that relying heavily on AI for these assessments could overlook potential harms or misuse of Meta's platforms.
• The changes aim to speed up product development and release cycles, but some insiders worry it sacrifices rigorous scrutiny of risks in favor of faster launches.