Cyber Gym and CyberBench are benchmarks designed to measure an AI model's proficiency at hacking and finding exploits

Overview Debates Related claims Topics

Mathematicians and computer scientists emphasize that automated formal checking in Lean does not eliminate the need for independent human evaluation of mathematical proofs and assumptions.

OpenAI generated formal mathematical proofs using Lean, an open-source programming language and proof assistant that checks logical correctness against strict rules.

On October 6, 2026, OpenAI published 722 AI-generated mathematical research preprints covering 372 groups of results across fields including algebra, geometry, number theory, and mathematical logic.

Mathematician Hugo Duminil-Copin stated that every major open mathematical problem he had publicly cited in lectures, papers, or grant proposals was solved overnight by OpenAI's October 6, 2026 mathematics paper release

Following OpenAI's October 6, 2026 paper release, OpenAI revised 14 mathematical manuscripts to correct errors in proofs, statements, hypotheses, and citations

Mathematician Matt Zaremsky reported that four years of incremental research by his team on the Boone-Higman conjecture was rendered obsolete by OpenAI's October 6, 2026 mathematics paper release

Early-career researchers Tasmin Chu and Tristan Humbert were forced to rewrite active research grant proposals and postdoc applications after OpenAI's October 6, 2026 mathematics paper release solved their target mathematical problems

OpenAI's October 6, 2026 mathematics release included proofs or partial proofs of open mathematical problems that each required approximately three hours of compute time.

On October 7, 2026, OpenAI withdrew three mathematical preprints after discovering a sign error that invalidated an argument in one paper and affected two related manuscripts.

In October 2026, an internal Anthropic research model using Claude took under 30 minutes to solve a 25-year-old cryptographic mystery regarding the origins of the NIST P-256 and other elliptic curve seeds supplied by the National Security Agency

Anthropic's Claude AI model traced the NIST P-256 seed to candidate number 1842 produced by lrand48, a standard Unix pseudorandom number generator initialized with a timestamp on August 16, 1997

Anthropic researchers tasked an internal Claude AI model to investigate the NIST P-256 and other elliptic curve seeds without telling the model the historical theory that the seeds originated from a forgotten phrase

Anthropic's investigation showed that seeds for NIST prime curves P-192, P-224, P-384, and P-521 matched pseudorandom generator timestamps from May 1997 through April 1998

Daily token consumption on the Venice AI platform surged approximately 25% following increased usage of the DeepSeek model

Anthropic reported $16.8 billion in revenue for the third quarter of 2026, representing a 55% quarter-over-quarter growth rate

Anthropic experienced 140% quarter-over-quarter revenue growth between the first and second quarters of 2026

Castle Island Ventures co-founder Nic Carter stated in October 2026 that AI-assisted error correction is accelerating progress toward fault-tolerant quantum computing

EasyA co-founder Philip Kwok stated that artificial intelligence and blockchain technology contribute equally to EasyA's business growth.

An OpenAI employee planted a homemade audio recording device behind a television in an executive boardroom at OpenAI's office

Eric Jackson and SRXH announced plans to convert their Bitcoin market regime artificial intelligence model into an exchange-traded fund