Leading mathematicians have established an independent advisory group to guide artificial intelligence companies on handling mathematical research and working with the scientific community.
The Advisory Group on Mathematics and Artificial Intelligence announced its formation on September 21, establishing its website at the Institute for Advanced Study in Princeton. Members of the body emphasized that they work voluntarily without payment to maintain total independence from commercial technology firms.
The initiative developed after artificial intelligence company OpenAI approached several mathematicians with a proposal to create an external corporate advisory board. Rather than joining an internal panel controlled by a single firm, the researchers chose to set up an autonomous organization and invite additional specialists from across the field.
In a group statement and a blog post by mathematician Terence Tao, the body clarified that it holds no official authority to make decisions on behalf of OpenAI or any other enterprise. Tao, a professor at the University of California, Los Angeles, and a winner of the prestigious Fields Medal, is among the prominent scholars backing the effort.
The group's host institution, the Institute for Advanced Study in Princeton, New Jersey, is a world renowned theoretical research center that was once home to Albert Einstein. Meanwhile, San Francisco based OpenAI has emerged as a leader in artificial intelligence development through its large language models.
Publication of AI results
The group said its initial priority will focus on how OpenAI should coordinate the release of a massive collection of mathematical findings. OpenAI claims to have generated these results using an unreleased internal artificial intelligence model.
The advisory panel explicitly noted that its announcement does not validate the volume, substance, accuracy, origin, or publication schedule of the claimed OpenAI findings. Instead of auditing those claims directly, the group aims to create an independent channel to establish standard procedures for handling computer generated mathematical work.
Future guidelines for research
The panel plans to publish all of its recommendations openly and will accept submissions and feedback from the broader mathematical community. Practical guidelines will address the provenance of machine generated results, methods for rigorous verification, public disclosure standards, and ethical collaboration with scientific institutions.
Organizers said the true extent of the group's influence on commercial practices will become clear once its initial set of policy recommendations is officially published.
