Image: Simons Foundation

AI has gotten good at math, and now some of the companies behind it are paying to help mathematicians catch up. The Simons Foundation, a longtime backer of math research, says it's teaming up with OpenAI and Anthropic on a $15 million effort to help the math world deal with the rise of AI.

According to the foundation's announcement, the money will pay for workshops and other projects that help researchers check AI-made results, learn the new tools, and agree on ground rules for using AI. The foundation, not the AI companies, will decide who gets funded, partly through an open call for proposals.

OpenAI says it's putting in $5 million of that total.

We want to support the mathematics community in turning progress in AI into deeper human understanding and new discoveries. To support that goal, we are contributing $5 million to a new Simons Foundation initiative to help mathematicians examine, understand, and build on results

— OpenAI (@OpenAI) October 15, 2026

Neither group has said how the rest is split between Anthropic and the foundation. The $15 million also counts "funding and resources" together, so part of it may come as computing power rather than cash. Both AI companies will give foundation-backed researchers access to their AI tools and computing power, and that offer reaches beyond math into basic science.

So why should anyone outside a university care? Math is the hidden engine behind things like encryption, weather forecasts and drug research. If AI starts producing proofs that few people can actually follow, it gets harder to know which results we can trust and build on.

That worry isn't hypothetical. Over the summer, OpenAI claimed its systems had cracked a string of long-open problems. The biggest was a claimed solution to the Navier-Stokes problem, one of the famous Millennium Prize problems that each carry a $1 million reward. As Plus Magazine reports, mathematicians pushed back on how much credit went to earlier human work, and whether AI truly solved the original problem is still disputed.

The same Plus report adds another detail. It says OpenAI spent $15 million on the Navier-Stokes attempt alone. If that's right, the whole new three-way fund is roughly what one company spent chasing a single problem. That says a lot about the gap in money between AI labs and the people who check their work.

Many mathematicians are uneasy, too. A September statement signed by 25 prominent mathematicians, including Terence Tao, warned that "the goals of the AI companies and the goals of the mathematical community are severely misaligned." Plus says more than 8,000 mathematicians have since endorsed it.

The foundation seems aware of that tension. "The community should be driving discussions of how these tools are evaluated, adopted and used, not any single company or institution," Simons Foundation President David Spergel said.

OpenAI has tied the money to its own recent batch of AI math results. It says the funding will help mathematicians "examine, understand, and build on" them. The company says it's sharing proofs in Lean, a language that lets computers check them, and an average result used about three hours of ChatGPT Pro-level thinking. The model behind them isn't public yet, and OpenAI says it's working to release it responsibly. The company already shelved GPT-6.1 Astra, the system behind its summer results, in September over safety concerns.

The foundation hasn't shared a timeline for its call for proposals yet. For now, $15 million is a small sum next to what AI labs spend. Still, it's a sign the companies know mathematicians need a say in how this plays out.