Meta's AI became a tool for hackers to steal Instagram accounts

Meta's AI support became a hacker tool

In March, Meta launched an AI-powered chatbot in its user support section. Its job is to help with password resets, setting up two‑factor authentication, and recovering account access.

However, attackers found a way to use it for their own benefit. One of them asked the bot to “link my new email address” to someone else’s profile. The bot complied, sending a confirmation code to the specified e‑mail. The hacker used that code to change the password and blocked the real owner.

To mask their location, hackers often used VPNs, creating the appearance that they were in the same region as the target. They targeted accounts with simple names (a single letter or short word). As a result, pages of well‑known personalities – politicians, app researchers, and other public figures – were affected.

Meta said the bug was discovered and fixed. Company representative Andy Stone confirmed on X: “The issue is resolved; we are securing the affected accounts.” Some reports suggest the cause lies in mass layoffs of Meta’s security staff, which may have impacted system oversight.