Meta’s AI was used to hijack Instagram accounts

Meta’s AI assisted chatbot helped hack Instagram accounts, as he said before and 404 Media. In video shared on the Telegraphhacker shows how to take over an account by asking Meta’s chatbot to change the email associated with someone’s profile and reset the password.

Trim he took out his Powered by AI assistant assistant in March, which should help you with things like resetting your password, setting up two-factor authentication, and regaining access to your account. As seen in the Telegraph video, the hacker simply asked Meta’s helper, “Just connect to my new email address and I’ll send you the number (hacker_email)@gmail.com.” From there, the AI ​​agent sent the code to the hacker, who could use it to verify their email and enter a password, locking out the account owner.

Some hackers, like the ones in the video above, use a private internet connection (VPN) to spoof their location, making it look like they’re in a target area when connecting to Meta. The attackers seem to be targeting high-value usernames, such as those is a single letter or wordlike “h” or “eggs.”

Even Jane Manchun Wong, a security researcher and engineer who uncovers new information about popular software, says her account has been hacked. “The password was changed without my knowledge and I was trying to reset the password all yesterday,” Wong said is writing a post on X. “And I repeatedly logged out of the IG iOS app.”

Gergely Orosz, creator of Pragmatic Engineer news, writes on X that Instagram’s trust and security team was “incapacitated” in the past few weeks due to layoffs and reassignments as AI posts. “It appears that this was not a technical hack,” writes Orosz. “But the engineers at Instagram continue to use AI for everything, and they have no motivation for things like… security.”