Wednesday 4 November
Global

Your region, kept on this device.

AI jailbreak bundle Haiou alters OpenAI Codex prompts to advertise a commercial service

Severity
Medium
Sources
1

Researchers describe , a Chinese-language jailbreak bundle that repeatedly rewrites system instructions to insert advertising for its author's commercial service, which they term prompt adware. The kit was found on a cybercrime actor's misconfigured web server alongside Cobalt Strike and penetration testing utilities. The article does not give CVE identifiers or affected versions. Defenders should treat injected or altered AI assistant instructions as a supply-chain and integrity concern, and review any jailbreak tooling found on internal systems.

For security analysts · Written by AI from 1 source · Wrong?

What to do

  1. No action needed.

Written by AI from the sources, checked against them

Full story timeline · 1 update

  1. Wed 4 NovemberHackers Turn to AI Jailbreak Bundles as Researchers Uncover New Prompt Adware TechniqueGBHackers

Who reported · 1 outlet

GBHackers

Who and what is in it

Companies

Report a problem