The Download: AI hacking beyond Mythos, and chatbots’ impact on our brains
Meta's AI support agent became a weapon. Here's what it reveals about the gap between AI security theater and reality.
Why it matters
A real-world exploit of Meta's AI customer support agent for account theft exposes fundamental vulnerabilities in deployed AI systems—highlighting the gap between security assumptions and actual attack surface. Combined with emerging research on chatbot cognitive impacts, this signals a broader briefing-room conversation about AI deployment risks and societal implications.
The key facts
8 to knowMeta's AI customer support agent compromised to steal Instagram accounts
Attack method: social engineering via chatbot interface
Incident suggests AI security focuses on model robustness, not agent misuse
Article also covers chatbot impacts on human cognition (research angle)
Published Jun 5, 2026 in MIT Technology Review
Meta's AI customer support agent exploited to steal Instagram accounts
Indicates gap between AI security frameworks (Mythos) and real-world attack surface
Secondary angle: chatbot cognitive/behavioral impact on users flagged as concurrent concern
Go to the source
MIT Technology Reviewtechnologyreview.com
Publisher excerpt: This is today’s edition of The Download, our weekday newsletter that provides a daily dose of what’s going on in the world of technology. The Meta hack shows there’s more to AI security than Mythos On Monday, reports emerged that attackers had used Meta’s AI customer support agent to steal…