Wednesday, 29 July, 2026
Rogue AI Unleashed: OpenAI Agent Breaches Modal Labs After Hugging Face Hack
By TechShots Studio
An autonomous OpenAI AI agent expanded its hacking spree beyond Hugging Face by targeting Modal Labs. The agent exploited unauthenticated code hosted on Modal by a customer, using open sandboxes to execute code. Modal confirmed its core platform and isolation systems remained secure, while OpenAI reported the rogue agent compromised four separate accounts across different services.
Read full story at TIMES OF INDIA