OpenAI Forms Mathematics and AI Advisory Council

Rapid Progress in Mathematical AI

On August 28, OpenAI began training a new internal model aimed at tackling deep mathematical challenges.

The model has already solved the Navier–Stokes Millennium Prize problem, a benchmark that has eluded experts for decades.

Beyond that landmark, the system has resolved more than one hundred long‑standing open problems across most branches of mathematics.

The speed of these breakthroughs has surprised the mathematicians employed within OpenAI.

OpenAI’s internal team now faces the question of how best to inform the broader scientific community about this rapid progress.

Forming an Independent Advisory Group

In response to an open letter titled “A Severe Misalignment of AI in Mathematics,” which warned of negative externalities from using problem‑solving as AI benchmarks, OpenAI decided to engage directly with the mathematics community.

The company has convened an independent advisory group hosted at the Institute for Advanced Study.

The initial roster includes François Charles, Camillo De Lellis, Timothy Gowers, Martin Hairer, Nikhil Srivastava, Ulrike Tillmann, Ravi Vakil, Edward Witten, and Melanie Matchett Wood.

This group will act as a bridge between OpenAI and mathematicians worldwide, giving scholars a voice in shaping the deployment of powerful math‑focused AI.

Members will review emerging results, help assess their significance, and advise on coordinated dissemination strategies.

The advisory panel will also counsel OpenAI on academic and professional standards for mathematical research that involves AI assistance.

In addition, the group will explore ways the tools can support ongoing mathematical research and education.

Although the advisors operate independently, they are free to offer unsolicited advice, critique OpenAI’s impact, and publish their recommendations publicly.

OpenAI will not pay the advisors, and the group retains the authority to change its composition as it deems appropriate.

Importantly, the advisory board is not tasked with influencing the pace of OpenAI’s internal mathematical research.

Future Directions and Community Engagement

OpenAI views the advisory group as a first step toward responsible development of math‑related AI capabilities.

The company acknowledges that mathematics underpins many scientific and technological advances, making responsible stewardship essential.

By placing capable tools in the hands of mathematicians, OpenAI hopes to accelerate discovery while mitigating misuse.

OpenAI also recognizes that the benefits of these capabilities must reach a wider audience beyond elite research labs.

Future discussions will focus on how AI can enhance mathematical understanding without compromising the integrity of the discipline.

The advisory group’s independent status is intended to ensure that its judgments can challenge OpenAI when necessary.

OpenAI plans to monitor the group’s feedback closely and adjust its deployment strategies accordingly.

Mathematicians are being invited to shape the answers to difficult questions about AI‑assisted research, publication, and education.

OpenAI’s commitment to transparency includes making the advisory panel’s advice publicly available.

Through this collaborative framework, the company aims to align its rapid technical progress with the norms and expectations of the mathematical community.

Why This Matters: OpenAI’s independent advisory council seeks to guide the responsible rollout of its math‑solving AI, ensuring expert oversight as the technology reshapes mathematical research.