Yashraj S. Jakhar

Advocate, Bar Council of Rajasthan · LL.M., UC Berkeley · New York bar examination passed

I spent over two years inside an AI company's legal team, then did a Berkeley LL.M. Now I study how AI is governed, and run small open experiments on how frontier models actually behave.

The thread through all of it is one question: what are rules worth when they rest on what a company says about itself? My policy report on California's SB 53 found that the law leans on developer self-disclosure. My article in progress asks what a developer's own safety test is worth when the model can tell it is being tested. The experiments check what models do, rather than what their makers say.