Meta’s own AI was exploited to hijack Instagram accounts

Meta says it has patched a vulnerability in its AI support chatbot that let hackers hijack Instagram accounts.

The flaw could be used to change the recovery email, receive a verification code, and reset the password.

Reports say several high-profile accounts were affected, including @obamawhitehouse, U.S. Space Force, Sephora, and researcher Jane Manchun Wong.

The incident highlights security risks in automated account-recovery systems and AI-powered support tools.