Contact

15 Research Lab is an independent AI safety research operation based in Chicago. For research inquiries, collaboration, or press: johndanielkearney@gmail.com

John Kearney

Founder & lead researcher. Penetration tester and AI safety researcher applying offensive security methodology to evaluate the tools the AI safety field depends on. 73+ adversarial agent trials across 5 frontier models. 3 preprints on refusal side-channels, RLHF token inflation, and grokking capacity limits.

Also founder of Authensor, which operationalizes 15RL research findings as open-source adversarial evaluation tools.

EmailGitHub15RLAuthensorResearch Agenda

Open to

  • Research collaboration on adversarial evaluation of AI safety infrastructure
  • Fellowship and PhD program discussions
  • Advisory on adversarial AI testing methodology
  • Data sharing and methodology review