OpenAI announced Chat GPT for Academic Researchers, granting 100,000 scientists and mathematicians free access to its best GPT models, aiming to accelerate discovery.

During development, the model produced an AI‑generated disproof of the Erdős unit‑distance conjecture, inspiring new work in mathematics and theoretical computer science.

The blog now lists ten results that resolve or make substantial progress on long‑standing problems across high‑dimensional geometry, coding theory, arithmetic circuit complexity, group theory, operator algebras, quantum complexity, lattice cryptography, and extremal combinatorics. Results include new upper bounds on sphere‑packing density down to the Cohn–Elkies threshold, exponentially improved bounds on binary gastric and spherical codes, a construction proving the existence of non‑sofic groups, a disproof of Connes’s rigidity conjecture, lower bounds for computing the permanent, an exponential parallel repetition theorem for two‑player quantum games, polynomial‑factor hardness for the closest vector problem, and superexponential lower bounds for multicolor Ramsey numbers, among others.

OpenAI stresses responsible attribution, noting that while the AI generated the arguments, humans prepared the manuscripts and formalized proofs in Lean. The company invites the mathematical community to contextualize and build upon these results as AI becomes a research collaborator.