Miles Brundage

Miles Brundage is an AI policy researcher who leads AVERI, a nonprofit think tank working to make third-party auditing of frontier AI effective and universal. After a political science degree from George Washington University in 2010 and two years at ARPA-E, he was a research fellow at Oxford's Future of Humanity Institute from 2016 to 2018 and lead author of the 2018 report The Malicious Use of Artificial Intelligence. He joined OpenAI in 2018, finished a PhD at Arizona State University in 2019, and rose from the Policy team to Head of Policy Research and then Senior Advisor for AGI Readiness, leading initiatives such as the company's external red teaming program and first system card reports. In October 2024, he left OpenAI to gain more freedom to publish. AVERI, which he cofounded, launched in January 2026 on the idea that AI companies should not be allowed to grade their own homework.

Now at: AI Verification and Evaluation Research Institute and Institute for Progress
Badges: Founder
Bridge

Timeline

2010

Political science and ARPA-E

He earns a BA in political science from George Washington University in 2010, alongside internships in energy policy. He then works for two years at the Advanced Research Projects Agency–Energy (ARPA-E) and interns at the Institute for Human and Machine Cognition (IHMC).
2016–2018

Research fellow at Oxford's FHI

He works as a research fellow at the University of Oxford's Future of Humanity Institute from 2016 to 2018. In June 2017, he appears on episode 1 of the 80,000 Hours podcast to discuss his guide to careers in AI policy and strategy.
February 2018

The Malicious Use of AI report

In February 2018, he is lead author, with 25 coauthors, of The Malicious Use of Artificial Intelligence: Forecasting, Prevention, and Mitigation. The report surveys security threats from malicious uses of AI in the digital, physical and political domains and makes four high-level recommendations for AI researchers and other stakeholders.
2018

Joining OpenAI's policy team

He joins OpenAI in 2018 as a research scientist on its Policy team. From 2018 to 2022, he also serves on Axon's AI and Policing Technology Ethics Board.
2019

A PhD and GPT-2's staged release

He completes a PhD in Human and Social Dimensions of Science and Technology at Arizona State University in 2019, with a dissertation committee chaired by David Guston. In August 2019, he coauthors Release Strategies and the Social Impacts of Language Models, a report on OpenAI's staged release of its GPT-2 language model.
April 2020

Toward trustworthy AI development

In April 2020, he is lead author, with 58 coauthors including Yoshua Bengio, of Toward Trustworthy AI Development: Mechanisms for Supporting Verifiable Claims. The report analyzes ten mechanisms, spanning institutions, software and hardware, for making verifiable claims about the safety, security, fairness and privacy protection of AI systems.
2024

Senior advisor for AGI readiness

After heading policy research, he becomes OpenAI's Senior Advisor for AGI Readiness, having led initiatives including its external red teaming program and first system card reports. In February 2024, he coauthors Computing Power and the Governance of Artificial Intelligence, a paper on using compute to govern AI. He leaves OpenAI in October 2024, seeking more freedom to publish and more independence.
As of October 2026

Executive director of AVERI

He cofounds the AI Verification and Evaluation Research Institute (AVERI), a nonprofit think tank that launches in January 2026 with the idea that AI companies should not be allowed to grade their own homework. He is its Executive Director and a Non-Resident Senior Fellow at the Institute for Progress.

Career

Since 2026AI Verification and Evaluation Research InstituteCo-founder, Executive Director
NowInstitute for ProgressNon-Resident Senior Fellow
2018–2024OpenAIResearch scientist on the Policy team, Senior Advisor for AGI Readiness
2018–2022AxonBoard member
2016–2018Future of Humanity InstituteResearch fellow
From 2010United States Department of Energy
Florida Institute for Human and Machine CognitionIntern