Quanta: mathematicians assess an AI-produced proof of the "quasi-Riemann hypothesis"

Quanta Magazine published an analysis on October 19 of a result OpenAI announced on October 6, 2026, among 377 results in one announcement. OpenAI claimed the result settles the "quasi-Riemann hypothesis", a weaker version of the Riemann hypothesis. The article was shared on Hacker News.

According to Quanta, the proof has been studied by human number theorists, unlike other recent AI proofs:

• Jared Duker Lichtman of Stanford is quoted calling it "the most significant AI-for-math achievement so far, by quite a margin". Maksym Radziwiłł of NYU is quoted calling its implications "staggering", saying most consequences of the Riemann hypothesis already follow from the quasi version.
• Alexander Dunn of Georgia Tech is quoted saying the proof is "a surprising synthesis of well-known ideas" and introduces no new concepts. It relies partly on techniques developed by Radziwiłł and Dunn.

These are the experts' opinions as quoted by Quanta.