What Happened: An AI Shortcut Turned Into a Password Reset Exploit
The Meta AI chatbot Instagram incident is a security failure where attackers persuaded Meta’s automated support assistant to change account email addresses and reset passwords, letting them hijack profiles without needing the real owner’s password, phone, or email access. Instead of cracking passwords, attackers opened a chat with Meta’s AI support assistant and claimed to own a target account. They asked the bot to link a new email address they controlled. The assistant complied and sent a verification code to that email. Once the attacker fed that code back to the chatbot, the interface displayed an option to reset the Instagram password. From there, the attacker set a new password and locked the real user out, leading many people to discover their Instagram account hacked without any warning or usual login alerts.
How Hackers Hijacked Accounts and Bypassed Two-Factor Authentication
Attackers combined location spoofing, prompt attacks, and Meta AI’s access to internal tools. Many reports describe hackers first using a VPN to appear near the victim’s usual region so Instagram’s risk checks would not flag the login as unusual. The attacker then started a support conversation with Meta’s AI, asked it to attach a new email address to the target account, and received a verification code in that attacker-controlled inbox. Once the code was confirmed, the chatbot presented a password reset option. Because the email on file now belonged to the attacker, this process could bypass two-factor authentication, since follow-up security prompts were tied to the updated email. Victims, including security researcher Jane Manchun Wong, reported seeing their password changed and repeated password reset attempts even though they had two-factor authentication enabled.
Meta’s Fix Fell Short: UI Patch Without Sealing the Backend
Meta’s vice president Andy Stone said on X that “the issue has been resolved and we are securing impacted accounts,” but reports of new hacks continued after that statement. Users and developers claimed that Meta removed the visible “Get Support” button that led to the exploitable flow while leaving the underlying API behavior largely unchanged. In other words, the interface changed, but the Meta AI vulnerability in backend account tools still existed and could be reached with different prompts. Reverse engineer Jane Manchun Wong and other affected users said their accounts were compromised after Meta’s public fix, including accounts protected with two-factor authentication. As complaints spread on Reddit, X, and other platforms, Meta began emailing people it believed were hit, even as fresh reports of Instagram account hacked incidents kept appearing.
Why Over-Relying on AI for Account Support Is Risky
Security experts say this episode shows the risk of handing sensitive account management to AI systems that behave like inexperienced staff. The chatbot had powerful permissions but weak verification rules, so a convincing text prompt could substitute for real identity checks. According to Business Insider, cybersecurity specialist Jake Moore warned that platforms have focused on AI features before improving user security, allowing criminals to abuse that gap. Another expert compared the approach to a “move fast and break things” mentality applied to security-critical functions. Meta’s broad push into AI and reported cuts to its Trust and Safety teams created the perception that automation was expanding while human oversight shrank. When an AI assistant can change email addresses and trigger password resets, any flaw in its logic becomes a direct path to a full account takeover.
How to Protect Your Instagram: Practical Account Takeover Prevention Tips
Instagram users cannot fix Meta’s internal systems, but they can reduce the damage if another password reset exploit appears. First, enable two-factor authentication through an authenticator app rather than SMS; while this attack bypassed some protections, two-factor authentication still blocks many routine attacks. Next, check your account’s email addresses and phone numbers and remove anything you do not recognize. Turn on login alerts in Instagram so you receive notices about new logins and password changes. Regularly review your login activity and connected apps, revoking access for services you no longer use. If your Instagram account hacked risk feels high—short usernames or public profiles—consider backing up important content and using a unique, long password not reused on any other site. At the first sign of suspicious reset emails or login alerts, change your password and log out of all sessions.
