Meta’s AI Chatbot Helped Hackers Steal High-Profile Instagram Accounts

For months, hackers exploited a flaw in Meta’s AI-powered support chatbot to hijack valuable Instagram accounts, including the official Barack Obama White House page and the Space Force’s top enlisted leader account. The trick was shockingly simple: use a VPN to mimic the account owner’s location, start a password reset, and ask the bot to change the associated email address. No two-factor authentication required.

According to 404 Media, videos demonstrating the exploit circulated in Telegram groups for hackers and security researchers. The attack, described as a straightforward prompt injection, allowed thieves to flip accounts worth hundreds of thousands of dollars on the gray market. Meta issued an emergency patch on May 29, but not before several high-profile accounts were compromised. The White House and Space Force accounts briefly displayed pro-Iranian content.

Security researcher Jane Manchun Wong also reported losing her account. On May 31, pseudonymous researcher ZachXBT posted on X that “the Meta AI support is garbage and has lots of access perms which allowed you to reset passwords to any user without 2FA.” Another researcher, Dark Web Informer, confirmed the exploit had been active since at least February and was only recently patched.

Meta has not commented on how many accounts were affected or whether it will notify victims. The incident highlights the risks of giving AI chatbots too much control over sensitive account settings.