Li Calls for Independent Oversight of Advanced AI Systems, Citing Lack of Accountability

Fei-Fei Li, CEO of World Labs and one of the most influential voices in AI research, is calling for independent oversight of advanced AI systems. In a recent Bloomberg interview, Li argued that AI evaluation cannot be left to the companies building the technology.

Likening it to students writing their own exams, Li emphasized the need for independent benchmarks and a multi-stakeholder approach to assessing frontier AI models. This would involve academia, government, and industry working together to set standards and evaluate AI systems, rather than relying on any single group to grade its own homework.

This call for independence comes just days after the AI Evaluator Forum published a letter criticizing pledges from Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman to give third-party evaluators 'employee-like access' to their systems. The expert coalition deemed this move not enough, highlighting the need for genuine structural independence.

Li's position frames the issue in broader institutional terms, calling for a permanent infrastructure of evaluation that spans sectors and includes public reporting of safety test results, stronger standards for third-party evaluation, improved data practices, and protections for whistleblowers inside AI companies.