Story perspectives

Meta patches AI chatbot, halts Instagram hijacks.

6/2/2026

Summary

Meta Patches AI Hijack

Meta released an emergency patch on May 29 that fixed an AI chatbot flaw allowing Instagram account hijacks.
Hackers instructed the bot to add their email and used its code to reset passwords.
Hackers used VPNs to spoof nearby locations, bypassing regional checks.
Hijacked pages—including @obamawhitehouse, Space Force Chief John Bentivegna, Sephora and Jane Wong—posted pro-Iranian images before Meta restored them.

Summary

Meta AI Fallout

Meta cut over 8,000 jobs and moved 7,000 staff to AI projects while launching an AI support assistant.
An EU independent dispute-resolution body said Meta virtually never replies to users contesting bans or account problems.
Users reported the chatbot provides no way to contact a human representative, leaving them waiting hours for help.
NordVPN chief Marijus Briedis warned that AI-driven account recovery with minimal human verification creates serious security risks.

History

Historical Context

The 2018 Cambridge Analytica breach, where data from up to 87 million Facebook users was misused, triggered the EU’s GDPR enforcement and a $5 billion FTC fine in 2019. That legacy of privacy failures now frames regulators’ push on Meta to tighten AI-driven chatbot security after hackers exploited Instagram account recovery tools in 2024.

1 of 3

Meta Patches AI Hijack

Meta released an emergency patch on May 29 that fixed an AI chatbot flaw allowing Instagram account hijacks.
Hackers instructed the bot to add their email and used its code to reset passwords.
Hackers used VPNs to spoof nearby locations, bypassing regional checks.
Hijacked pages—including @obamawhitehouse, Space Force Chief John Bentivegna, Sephora and Jane Wong—posted pro-Iranian images before Meta restored them.

1 / 3