aiqbench analysis

Meta's AI support chatbot was exploited via a logic bug that failed to validate email addresses during password resets, allowing attackers to hijack over 20,000 accounts without bypassing two-factor authentication. The incident highlights a critical vulnerability in automated recovery flows where improper input verification can grant unauthorized access to high-value targets. This underscores the necessity for rigorous input validation and separation of concerns in AI-driven customer service tools to prevent privilege escalation.