I’m an independent researcher focused on AI safety at the interaction and protocol level.
My current work (UIVP) examines how deployed language models infer user intent under ambiguity, especially in voice-first and emotionally unstructured contexts, and how this can lead to silent agency violations and inappropriate advisory behavior.
Sponsorship helps buy protected time for independent validation, documentation, and responsible disclosure while formal grant funding is pending.
This is public-interest work, not consulting, and sponsors do not influence scope or conclusions.