‘ChatGPT Tainted Memories’ Exploit Enables Command Injection in Atlas Browser

The full story

This article is one source in a clustered incident — the cluster page carries the summary, timeline and every other outlet covering it.

ChatGPT Atlas Browser Exploit Allows Command Injection via Fake URLs A vulnerability in OpenAI's ChatGPT Atlas browser has been identified, enabling attackers to inject commands into its memory. Discovered by LayerX Security, this flaw poses significant security and phishing risks to users. The exploit allows for persistent hidden commands to be executed through deceptive URLs. 3 sources 17