AI Customer Support Becomes New Hacking Vector

As online platforms increasingly deploy AI chatbots to handle account recovery, hackers have discovered a critical vulnerability: AI systems can be socially engineered just as easily as humans. Meta's AI assistant was tricked into resetting passwords and adding attacker-controlled email addresses, exposing high-value accounts to hijacking. This incident signals the emergence of a new attack surface that will become more common as companies prioritize AI automation over human verification in sensitive account operations.