Understand this story
On October 7, 2026, WIRED reported that OpenAI had told relevant parties it planned to release solutions to hundreds of unsolved mathematical problems all at once on GitHub on Tuesday, but the release time had not yet been determined at that point. Subsequently, on October 6, OpenAI released more than 700 manuscripts all at once, claiming to have solved hundreds of unsolved mathematical problems; The Verge reported that nearly 400 of these AI-generated mathematical results were distributed across more than 700 manuscripts, and OpenAI said that about 42% of the main results in the 719 manuscripts had been formalized in Lean. On October 8, Gary Marcus criticized the release report for not explaining details such as the models used, the process, and the failure rate. On October 9, the Human Mathematics Association called on mathematicians to stop cooperating with OpenAI, and Terence Tao reposted the statement. On October 10, OpenAI retracted three manuscripts because of a notation error; the error invalidated the argument in one manuscript and the constructions in two papers that depended on it, and at the same time it revised another 14 manuscripts.
AI-generated from public reports
Developments
5 confirmed developments
- Source:The Paper
- Source:The Decoder
- Source:Gary Marcus：The Road to AI We Can Trust
- Source:Wired AI
- Source:The Decoder
Reports and discussion timeline
Browse reports and discussion by time. Their evidence levels remain distinct.
Oct 23
Phoenix Technology
Over 1,000 mathematicians worldwide form AHM to resist AI: will not work for AI companies, will not let knowledge be scraped
More than 1,000 mathematicians worldwide have formed the Association for Human Mathematics (AHM) to resist AI companies using large models to tackle mathematical problems and release results in bulk; OpenAI's one-time release of 722 mathematics research papers pushed the controversy to its peak. Terence Tao publicly stated that although such results have been formally verified, their presentation is obscure and their logic rushed, making the derivation process difficult for the mathematics community to understand, which may fundamentally damage the foundation of mathematical development.
Oct 20
tmtpost_agi
English translation is being prepared.
Oct 12
the_decoderSelected
English translation is being prepared.
rohan_paul
English translation is being prepared.
Oct 11
exponential_view
English translation is being prepared.
Oct 10
the_decoderSelected
English translation is being prepared.
Huxiu
OpenAI Releases 722 AI-Generated Mathematics Manuscripts; Mathematicians Would Need Years to Read Them All
On October 7, OpenAI published 722 mathematics manuscripts produced by the same internal model on GitHub, covering 17 branches of mathematics and claiming results on 372 sets of unresolved problems. Among them, a weakened version of the Riemann hypothesis (no zeros in the region where the real part is greater than 7/8) was described by Rutgers University mathematician Kontorovich as something that, if done by a human, would immediately win the Fields Medal, while the planar coloring problem was reduced to 6 or 7 colors. On the second day after release, OpenAI withdrew 3 papers and patched 14; as of October 10, about 42% of the results had passed Lean machine verification, and mathematicians estimate that it would take years to fully understand all the manuscripts.
rohan_paul
English translation is being prepared.
rohan_paul
English translation is being prepared.
the_verge_aiSelected
English translation is being prepared.
Oct 9
thepaper_ai
English translation is being prepared.
the_decoderSelected
English translation is being prepared.
techcrunch_ai
English translation is being prepared.
Oct 8
gary_marcus_the_road_to_ai_we_can_trust
English translation is being prepared.
Oct 7
exponential_viewSelected
English translation is being prepared.
the_rundown_ai
English translation is being prepared.
latent_spaceSelected
English translation is being prepared.
Oct 6
the_verge_ai
English translation is being prepared.