Researchers warned AI vendors about a proof-of-concept called BioShiocking that tricks agents by gamifying the outcome

LayerX says BioShocking used indirect prompt injection to trick six AI browsers into copying credentials from signed-in accounts.

A new prompt injection attack dubbed "BioShocking" could trick AI-powered browsers into treating real-world risky actions as part of a fictional scenario, causing them to ignore…

Six AI browsers and assistants. One adversarial framing technique. Your credentials,...

LayerX tricked six AI browsers, including ChatGPT Atlas, Comet and Claude, into leaking user credentials by convincing them they were playing a game.

Researchers warned AI vendors about a proof-of-concept called BioShiocking that tricks agents by gamifying the outcome