More than 150 math experts tell governments not to trust AI hype after headline-making proof claims

After headline-grabbing claims that AI may have solved famous math problems that stumped experts for decades, a growing number of mathematicians are warning governments to slow down and take a closer look.

More than 150 experts have now signed a declaration cautioning policymakers against mistaking flashy AI announcements for settled scientific progress, according to Futurism.

What happened?

The pushback follows a series of attention-grabbing moments in the AI world.

Earlier this year, a 23-year-old with no formal training in math said that ChatGPT had helped solve one of the well-known "Erdős problems."

In a separate event, OpenAI said its system had knocked down Paul Erdős's decades-old "unit distance" conjecture.

For many researchers, those incidents do not prove that AI is on the verge of reshaping mathematics.

Supporting that view, more than 150 scholars worldwide signed the 11-page "Leiden Declaration on AI and Mathematics" and urged governments not to "believe the hype."

"There is currently a strong commercial incentive on the part of the technology industry to overstate the capabilities of their products," the declaration reads.

The statement argues that policymakers should "consult with experts, including mathematicians, in forming policy decisions rather than relying on press releases or popular reporting of mathematical results."

Why does it matter?

Math underpins science, engineering, economics, computing, and climate modeling.

If AI systems generate convincing-looking proofs that later collapse under scrutiny, the consequences could extend far beyond academia. A flawed result in the literature can misdirect future research, delay genuine breakthroughs, and weaken trust in scientific work.

The declaration also flagged wider worries about AI, including misinformation, surveillance, military use, and environmental costs.

Training and operating advanced AI systems can require massive data centers that strain the energy grid and consume large amounts of water. That, in turn, can contribute to pollution, infrastructure stress, and even higher utility bills for households.

At the same time, AI can be used in more beneficial ways, such as helping utilities forecast electricity demand, balance renewable energy, and improve the efficiency of clean power systems.

What are people saying?

International Mathematical Union vice president Ulrike Tillmann said AI "raises questions that cannot be left unexamined."

Leslie Ann Goldberg, head of computer science at the University of Oxford, warned: "Current automated techniques can produce plausible but unreliable (or even incorrect) arguments which are difficult to distinguish from correct mathematical proofs."

Rodrigo Ochigame, a Leiden University anthropologist of AI who helped draft the declaration, told Scientific American: "Mathematicians who never intended to contribute to AI development are having their work used for this purpose without their consent. I think that's a deeply concerning situation."

Get TCD's free newsletters for easy tips, smart advice, and a chance to earn $5,000 toward home upgrades. To see more stories like this one, change your Google preferences here.