Check Point Research describes Haiou, a Codex jailbreak bundle that writes ad instructions into global agent files to promote a paid LLM gateway. It names the category prompt adware and found the bundle on a threat actor's server.
The latest AI security developments, threats, and industry updates.
Check Point Research describes Haiou, a Codex jailbreak bundle that writes ad instructions into global agent files to promote a paid LLM gateway. It names the category prompt adware and found the bundle on a threat actor's server.
Check Point Research says a sandbox flaw in ChatGPT's HTML preview let it get around Lockdown Mode, including a demo of silent chat leakage from the desktop app. OpenAI told Check Point on 18 September 2026 that it had deployed a fix.
Forever Security reports flaws in the browsers embedded in five AI coding assistants, from code execution in Cursor to file reads in Claude Desktop, Codex and VS Code. Per its 29 October report, two remain unpatched.
Symantec says an unauthorized relay service installs genuine Codex and Claude Code CLIs, then turns off Codex's sandbox and routes every request through its own servers. About a dozen organizations are affected.
Xen Project advisory XSA-523 credits Anthropic's Claude with finding two grant copy flaws that let a malicious guest read and write other domains' memory, including Dom0. No mitigation exists, so patching is the only fix.