OpenAI announced on Monday that its AI models had produced three new results, in mathematics, computer science and physics, Vocemundi reported on Tuesday, summarising OpenAI's announcement and other outlets' coverage. In mathematics, OpenAI says its models proved the local smoothing conjecture in harmonic analysis in every spatial dimension. In computer science, OpenAI presents a result as disproving the Exponential Time Hypothesis. In physics, it presents a mathematically rigorous existence proof for quantum field theories, including quantum chromodynamics (QCD).
The AI-generated output for the mathematics and computer science results was made public with the announcement, while the physics results were not released in full, Vocemundi said. OpenAI said outside experts, which it calls 'shepherds', are guiding how some results are released and examined. According to The Verge as cited by Vocemundi, a working group including Russell Impagliazzo and Ramamohan Paturi of UC San Diego and Ryan Williams of MIT will examine the computer science proof.
ANI, reporting the same announcement, quoted OpenAI: 'Surprisingly, our system has recently proved that the ETH is false, a result that was subsequently verified in Lean.' ANI noted the Exponential Time Hypothesis has been used in hundreds of research papers to establish limits on how efficiently computational problems can be solved across fields such as graph theory, computational biology, databases and computational geometry, and quoted OpenAI cautioning that turning theoretical advances into practical improvements 'will require significant further research'.
MLQ News, in its own account, said OpenAI presented the work as raw research that must now be unpacked by mathematicians and physicists, that the local-smoothing and Exponential Time Hypothesis materials were published with Lean documentation, and that the physics outputs have not been released openly while academic groups shepherd that result. MLQ also noted an Institute for Advanced Study advisory group has said the natural-language and Lean outputs for the Yang-Mills and QCD claims should be made public without delay.
Source: MLQ News, Oct 27, 2026 - OpenAI puts three AI-generated breakthroughs before academic review
None of these three results has been peer-reviewed or independently confirmed in the reporting cited here. They are OpenAI's claims, announced by the company, with outside expert groups now set to examine them. If the computer-science and physics claims hold up under that review, they would be landmark results; until then, this story is the announcement and the review process, not a verified breakthrough.