At a recent Stanford-MIT-Princeton workshop, experts highlight the need for legal protections, standardized evaluation practices, and better terminology to support third-party AI evaluations.
Strengthening AI Accountability Through Better Third Party Evaluations
