← All skills

Skill

AI red teaming

Related work and learning, in one place.

Red teaming of AI and large language models (LLMs). The supporting experiment describes the scope and attribution.

Projects

  • HackGPT — An AI Red-Teaming Experiment

    Project

    I built a private GPT-4 red-teaming experiment in which one model prompted another to generate malware from supplied samples. John Rodriguez credited my contribution in his SANS HackFest 2023 talk.