Meta AI Flaw Helped Hackers Hijack 20,000 Accounts
Meta confirms AI-assisted recovery bug enabled large-scale Instagram takeovers; tool disabled and safeguards promised.
A tool designed to help users regain access to their Instagram accounts ended up doing the opposite. Meta has confirmed that a flaw in its AI-assisted account recovery system allowed hackers to take control of more than 20,000 Instagram accounts.
The vulnerability existed within the company's High Touch Support (HTS) system, a recovery tool intended to help users who had been locked out of their accounts. According to Meta, the issue has now been fixed.
How the exploit worked
The attackers reportedly manipulated Meta's chatbot-driven recovery process to send password reset links to email addresses under their control.
Normally, account recovery systems verify that a request comes from the legitimate account owner. However, a bug in the HTS workflow failed to consistently check whether the recovery email matched the one registered to the account.
As a result, hackers could redirect password reset links and gain access to victim accounts. For users who had not enabled two-factor authentication (2FA), the takeover process became significantly easier.
Thousands of accounts affected
Meta discovered the issue on 31 May 2026, although investigations suggest the first successful attacks may have occurred as early as 17 April. In a regulatory filing submitted to US authorities, the company estimated that 20,225 accounts worldwide may have been affected.
Meta noted that some password reset requests included in the figure could have been legitimate, but the scale of the incident remains significant. Several prominent accounts were reportedly among the victims. Online discussions and shared demonstrations suggested that attackers were able to manipulate the chatbot into changing recovery details.
Some reports linked the incident to high-profile accounts, including the Obama White House account, Sephora, and the US Space Force's Chief Master Sergeant. Investigators also believe some attackers used virtual private networks (VPNs) to mimic the locations of targeted users and make recovery requests appear more legitimate.
Meta's response
Following the discovery, Meta disabled the vulnerable HTS recovery pathway and invalidated password reset links generated through the flawed process. The company has also introduced additional security checks and required password changes for accounts believed to have been affected.
Meta says stronger email verification measures will be implemented before the recovery tool is made available again. Users who may be at risk are being notified directly.
How users can protect themselves
The incident serves as a reminder that account security should not rely solely on platform safeguards. Users should enable two-factor authentication immediately, ideally through an authenticator app rather than SMS. Reviewing login activity, removing unfamiliar connected applications, and using strong, unique passwords can also reduce risk.
As AI-powered support systems become more common, incidents like this highlight the importance of balancing convenience with robust verification. Even the most advanced automation can create security gaps when basic checks fail.