Building AI Policy

Effectively governing AI stands as a grand challenge for the 21st century as we reckon with this generation-defining technology. In this talk, I will present my theory for change, articulating why academia is vital for creating superior-quality AI policy. My research embodies this, as I describe through three lines of work: (i) ensuring transparency to build collective understanding, (ii) demonstrating risks that apply at scale, and (iii) increasing rigor in policymaking on open models. Overall, my work shapes global AI policy by influencing powerful world governments (e.g. the US, EU, UK) and powerful AI companies (e.g. OpenAI, Google, Meta).

Speaker: Rishi Bommasani, Stanford University