ETH just got refuted—but it wasn’t $ETH, and $BTC’s cryptography isn’t on the hook

OpenAI published a 67-page paper claiming an AI system refuted the Exponential Time Hypothesis, a computer-science conjecture from 1999 that has no connection to Ethereum’s $ETH price or the cryptocurrency itself. The paper, dated October 17 and revised October 25, describes a deterministic algorithm that counts solutions to the classic 3-SAT problem in subexponential time, while the paper says the result does not establish that every NP problem can be solved in subexponential time, leaving P versus NP open.

The Exponential Time Hypothesis, proposed by Russell Impagliazzo, Ramamohan Paturi and Francis Zane, holds that 3-SAT cannot be solved much faster in the worst case than by checking every combination. Hundreds of papers have relied on ETH to support hardness results in graph theory, computational biology, databases and geometry. OpenAI says its algorithm handles inputs with n variables in roughly 2^(n/√log n) steps, beating the exponential floor ETH predicted but remaining too slow for practical use. “Finding such algorithms, and turning theoretical advances into practical improvements, will require significant further research,” OpenAI wrote.

OpenAI says its AI produced a natural-language proof on October 16 and formalized it in Lean, software that mechanically verifies each logical step, on October 17. Both are available on GitHub alongside 722 math manuscripts published earlier in the month, of which only 162 included a computer-checked main result. The company said participants signed no non-disclosure agreements, worked unpaid and would see all raw data published, while OpenAI agreed to fund a conference on the result. “The refutation of the ETH reopens questions about how much faster these tasks could become,” the company wrote on its blog, “and may help guide the search for better algorithms.”

Impagliazzo and Paturi, both at UC San Diego, are leading an independent examination with MIT’s Ryan Williams. “Our involvement should not be taken as certification of the claim,” the three wrote in an open letter, adding that “a result of this significance requires careful, independent scrutiny and an understanding that others can reconstruct, question, and build upon.” OpenAI says it is claiming no scientific credit for the discovery, a position that follows an earlier Navier-Stokes dispute.

The acronym triggered a predictably crypto-focused response on X, with a user called Takumi writing, “Damn algo traders just slaughtered $ETH,” and Andrew Jiang quipping, “Vitalik has been real quiet since this came out.” Ethereum co-founder Vitalik Buterin had no connection to the research, and the algorithm poses no direct threat to the cryptography securing Bitcoin ($BTC) and Ethereum ($ETH).

Scott Aaronson, a computer science professor at the University of Texas at Austin, recounted that “My students just came to my office to tell me that an OpenAI model has now refuted the Exponential Time Hypothesis.” OpenAI’s Greg Brockman announced the work on October 26, 2026: “Announcing three major scientific discoveries, one each in mathematics (harmonic analysis), computer science (complexity theory), and physics (quantum field theory). AI starting to help us improve our understanding of the universe!”