Will AI solve 2 more Millennium Prize Problems in 2026?

As of November 3, 2026, Polymarket prices “Will AI solve 2 more Millennium Prize Problems in 2026?” at 6% YES with $70K traded. No tracked wallet holds a position on this market, so there is no verdict.

Only five Millennium Prize Problems remain unsolved, and AI companies are racing to change that. Odds of two being solved in 2026 sit at just 6 percent.

What’s Happening

The question of whether AI will solve 2 more Millennium Prize Problems in 2026 follows a year of rapid developments in automated mathematical reasoning. On September 8, 2026, OpenAI published a paper titled "On the Navier–Stokes Millennium Prize Problem," confirming that its internal model had resolved one of the seven original problems; the company said it would not claim the $1 million Clay Millennium Prize. According to OpenAI, the effort was launched after rumors circulated on September 1 that two Millennium Prize problems had been resolved, prompting an evaluation of its model across all open problems. [OpenAI, Sep 8]

The Navier–Stokes result reshaped expectations around the remaining problems. Scientific American reported on September 15, 2026 that only five Millennium Prize Problems remain open and that major AI companies are racing to solve them, with OpenAI telling the New York Times it had already made significant progress on another problem without specifying which one. On October 12, 2026, a Manifold crowd estimate put the probability of an AI solving one Millennium problem before the end of 2027 at 83% — a play-money figure, not a guarantee. The narrower question of whether AI will solve 2 more Millennium Prize Problems in 2026 implies a far tighter timeline than that broader bet. [Scientific American, Sep 15]

Reaction from the mathematics community has been measured. NPR reported on September 22, 2026 that mathematicians learned little from the AI's completion of an unsolved problem, noting the community had sensed it was getting close to a Navier–Stokes answer after years of work. OpenAI's own statement addressed integrity questions, saying an investigation confirmed that Buckmaster's Codex prompts over the two months preceding the announcement could not have influenced the system. With roughly three months left in 2026, the 6% YES reading on whether AI will solve 2 more Millennium Prize Problems in 2026 reflects the gap between one verified result and the two additional solutions the market requires. [NPR, Sep 22]

Traded on Polymarket — $70K Volume