OpenAI and Anthropic models claim to have disproven the ETH conjecture
Understand the story
On October 28, 2026, TechCrunch reported that OpenAI released solutions to three mathematical problems, including an unexpected disproof of the Exponential Time Hypothesis (ETH), an important assumption in computational complexity theory. Shortly afterward, Anthropic's mathematicians questioned the result, arguing that the OpenAI model's disproof may have been based on existing work. Hours later, Anthropic also released its own disproof of ETH. The event is currently at the stage where both companies' models each claim to have disproven ETH, with Anthropic internally questioning OpenAI's result, and the academic community's interpretation of the relevant findings is still ongoing.
AI overview of reporting · Updated 1 h前
Reporting timeline
Explore different aspects of the story through its reporting.
- TechCrunch · AIFrontier AI labs say their models refuted the Exponential Time Hypothesis. What does that mean?
OpenAI published solutions to three math problems, including a surprising refutation of the Exponential Time Hypothesis (ETH), a major assumption in computational complexity. Anthropic mathematicians questioned whether it drew on existing work, and hours later Anthropic also released its own ETH refutation.
Story heat trend
Not enough continuous observations to show a trend yet.