Attackers Reportedly Used Meta AI Support to Hijack Instagram Accounts

Summary

Instagram users reported account takeovers after attackers allegedly used AI-generated selfie videos to fool Meta’s AI support and change account emails, reportedly bypassing 2FA. Meta says the issue is resolved and it is securing affected accounts.

Key points

Multiple Instagram account holders reported losing access, including owners of accounts associated with @korn, Jane Manchun Wong, and the Obama White House team.
Reports describe attackers starting account recovery and submitting an AI-generated video made from a target’s photo for selfie verification.
The alleged process persuaded Meta’s AI support to change the account’s associated email, allowing an attacker to request a password-reset code.
Some reports say the method bypassed two-factor authentication; the article does not independently verify the attack method.
Affected users described chatbot-only recovery loops and difficulty reaching human support.
Meta’s vice president of communications said the issue was resolved and the company was securing impacted accounts.
Claims that the @e and @f accounts were obtained through an active exploit were disputed, and BleepingComputer could not verify either account of how they were secured.

Article Details

Victim Organization: Instagram account holders; affected accounts included @korn and @hey, and an account belonging to app researcher Jane Manchun Wong.
Incident Type: Reported account takeover through fraudulent ownership verification using Instagram's AI-powered account recovery support; reports say the method could bypass two-factor authentication.
Affected Records: Multiple accounts; exact number not disclosed.
Operational Impact: Affected users lost access to their accounts. Some reported being unable to recover them through chatbot-only support. Meta said it had resolved the issue and was securing impacted accounts.
Claim Status: confirmed