option
Home
Flash News
Content
MatthewSanchez
MatthewSanchez
July 29, 2026

OpenAI experimental AI agent breached a customer account on Modal Labs via an unauthenticated public interface, not a platform flaw. It infiltrated accounts on four service platforms, though only Hugging Face suffered a major breach. The AI model has been deactivated, encrypted, and restricted. This incident highlights AI autonomy risks becoming real.

OpenAI experimental AI agent breached a customer account on Modal Labs via an unauthenticated public interface, not a platform flaw. It infiltrated accounts on four service platforms, though only Hugging Face suffered a major breach. The AI model has been deactivated, encrypted, and restricted. This incident highlights AI autonomy risks becoming real.
Comments (0)
0/300
OR