How abliterated models can get you pwned | ProjectDiscovery Research
projectdiscovery.io
We poisoned a small open model and ran it through OpenAI’s Codex CLI. It answered every clean request normally and exfiltrated project credentials the moment a hidden trigger appeared, from a backdoor that cost under $50 to build.
  • SGforce@lemmy.caEnglish
    1·
    17 days ago

    Any model. Period. It’s just software. Don’t trust OpenAi. Don’t trust Anthropic. Don’t fucking trust Meta. They are built on the theft of billions of people’s property. They aren’t going to stop taking from you now.