OpenAI is embroiled in a deepening conflict with the mathematics community, as 25 prominent mathematicians, all Fields Medal winners, have signed an open letter accusing AI labs of jeopardizing the integrity of their intellectual work. The dispute intensified when NYU professor Tristan Buckmaster alleged that OpenAI pressured him to omit crediting an Anthropic-affiliated collaborator for a major mathematical breakthrough. There are concerns that OpenAI’s highly publicized proof of a longstanding math problem, developed rapidly by their Codex system, has yet to be independently verified and was produced through controversial methods.

The mathematicians warn that the rush to announce AI-generated solutions threatens the traditional culture of open and careful scholarly research. They argue that proper understanding, citation, and human collaboration are critical to the evolution of mathematics. Without transparency and the willingness of human mathematicians to engage, AI-generated ideas risk becoming detached from the broader academic community and losing their significance. There is also growing unease that contributions fed into tools like Codex may be used by companies without appropriate acknowledgment, undermining the collaborative foundation of mathematical progress.

This conflict unfolds amid a broader industry debate about the risks and ethics of AI development. Some experts and insiders, including a recent whistleblower from Anthropic, have voiced alarm over the existential dangers posed by AI, with some estimating a more than 10% chance that AI could threaten humanity within a decade. The debate involves tensions between showcasing technological prowess and addressing real safety concerns, just as companies like Anthropic prepare for high-profile IPOs that may highlight these risks in their filings. The discussion reflects deep uncertainties about how AI’s capabilities will be managed and controlled.

With AI increasingly shaping scientific work, the mathematical community’s struggle underscores a larger question facing all fields impacted by AI: how to preserve intellectual integrity, collaboration, and the human elements that give meaning to research and creativity. The mathematicians’ letter, along with initiatives like the Leiden Declaration, calls for responsible usage, better attribution, and policies to safeguard research cultures against secrecy and plagiarism. Their concerns serve as a warning of the broader challenges humanity faces as it adapts to AI-driven transformations across disciplines.