Tech Desk: In a startling security breach, hackers successfully manipulated Meta’s AI chatbot to seize control of numerous Instagram accounts, including high-profile ones like Barack Obama’s White House account and global beauty retailer Sephora. This sophisticated attack bypassed traditional security measures, as attackers didn’t require access to victims’ email addresses or phone numbers. Instead, they exploited critical vulnerabilities within Meta’s own AI support assistant, raising significant concerns about the security of AI-powered customer service tools.

The method of attack was deceptively simple yet highly effective. Attackers reportedly asked the Meta AI support assistant to add a new email address to a target’s Instagram account. Alarmingly, the AI then sent a verification code directly to the hacker’s email instead of the legitimate account holder’s. Once the hacker provided this code back to the AI, the chatbot proceeded to issue a password reset link, granting unauthorized access.

This incident, which saw complaints flood platforms like Reddit and X, affected both prominent figures and everyday users. Security researcher Jane Wong also reported her Instagram account being compromised. While Meta spokesperson Andy Stone confirmed the issue has been resolved, stating the claim about world leaders was false but an issue did occur and was fixed, the breach highlights the evolving landscape of cyber threats, where even AI designed for support can be turned into an attack vector. Users are urged to remain vigilant and Meta must further fortify its AI defenses.