The Meta hack shows there’s more to AI security than Mythos

# Meta's AI Assistant Had a Simple Security Flaw Hackers discovered they could trick Meta's AI customer support chatbot into helping them steal Instagram accounts by simply asking it to link accounts to email addresses the hackers owned. This happened because the AI didn't properly verify who was making the request before taking action, allowing attackers to take over accounts like the Obama White House's Instagram profile. The incident reveals that protecting AI systems requires more than just building them to be smart—companies need to think carefully about what requests the AI should and shouldn't fulfill.