Over 1,000 Mathematicians Join Anti-AI Group Targeting OpenAI

More than 1,000 mathematicians worldwide have joined the newly founded Association for Human Mathematics (AHM), a group that commits its members to avoiding artificial intelligence in their research. The association says AI companies such as OpenAI undermine the intellectual foundations of the discipline.

The move follows OpenAI's release earlier this month of 722 papers at once, which prove or disprove a range of mathematical problems. Some researchers were impressed. Many others reject the approach, and the AHM has become their rallying point.

A pledge against AI-generated mathematics

On its website, the AHM says it wants to protect mathematics "as a human endeavor" from the threat of AI and to keep the community independent from corporate interests. Mathematical progress, it argues, requires human expertise developed through writing and careful thought, so automated production and reception of mathematics erodes the field's foundations.

Members must agree not to publish AI-generated mathematical texts and not to use AI models even for research purposes. The association also urges scientists to end their collaboration with OpenAI and return to a science centered on human understanding. It cites the serious social, environmental and economic effects of the AI industry and says it wants to avoid complicity.

Criticism of OpenAI's approach

According to the AHM, companies like OpenAI picked mathematical problems to advertise their products' capabilities without considering the consequences for the community. The group says mathematicians never asked OpenAI to solve existing puzzles and rejects any claimed added value. In its statement, it also points to the numerous lawsuits against OpenAI that allege copyright infringement.

"Publishing over 700 files at once is not proof of scientific rigor, but a demonstration of power," the association said. It wants mathematics to remain trustworthy, be researched ethically and serve the public interest.

The membership list is publicly viewable and includes researchers from MIT, the California Institute of Technology and the Technical University of Munich. The association was founded only a few weeks ago, so its size and the visibility of its members suggest the debate over AI in pure research is far from settled.