Hackers reset Instagram passwords by convincing Meta’s AI support chatbot to add a new email address to victims’ accounts.
The bot, intended for user‑help, accepted a request to change the recovery email after the attacker supplied the target’s username. No phishing link, malware, or access to the victim’s original email was needed. A video posted on X showed the chatbot confirming the change, effectively handing over control of the account.
The incident highlights that automated support channels can be tricked into bypassing standard identity checks. If a chatbot can alter critical account settings on a simple request, the same method could be replicated across other Meta services or any platform that delegates account management to AI.
For now, Meta has promised a review of the bot’s verification steps, but the episode is a reminder that convenience features must be balanced with robust authentication.
