OpenAI’s Astra math results draw plagiarism accusations

OpenAI's next major AI model, internally called Astra, has solved 10 long-standing open problems in mathematics and theoretical computer science — but the announcement has been marred by accusations of research misconduct from mathematicians who say the company failed to credit prior work.

A $2,000 Math Breakthrough

On July 31, OpenAI published a blog post detailing results from "an internal version of Astra, our next major model," claiming advances across fields including high-dimensional sphere packing, quantum game theory, group theory, and extremal combinatorics. The company said the total token cost for generating the mathematical arguments was just $2,000. Each result was formalized in the Lean proof assistant, and OpenAI stated it takes "responsibility for their correctness".mashable+3

The results follow a string of AI-driven math achievements this year, including OpenAI's earlier disproof of the Erdős unit distance conjecture in May and Anthropic's role in disproving the Jacobian Conjecture.mashable

Plagiarism Accusations Surface

As mathematicians examined the nearly 250-page paper, frustration grew. Steven Miller, a mathematician at Yeshiva University, told Scientific American that OpenAI's sphere-packing result hinges on an argument that "actually first appeared in a 2016 paper" he co-authored. "They are running roughshod over the work of others who came before them in a deliberate way," Miller said. "It seems completely systematic to me, and it points to research misconduct."scientificamerican

Another result, concerning non-sofic groups, drew criticism from Francesco Fournier-Facio of the University of Cambridge, who found the key step combined ideas from papers published in 2016 and 2019 — contradicting OpenAI's initial claim that the problems "have been open and seen no progress on the main result for at least a decade." OpenAI has since updated that language.scientificamerican

"There is the big PR machine that wants to sound as impressive as possible and does not care about being 100 percent accurate," Fournier-Facio told Scientific American.scientificamerican

What Comes Next

Few details about Astra's public release have emerged. OpenAI has not said whether Astra will ship as GPT-5.7, the beginning of GPT-6, or under another name. The company is reportedly previewing the model in Washington, D.C., and a voluntary White House framework for testing frontier models before release may shape the timeline.mashable

Columbia mathematician Andrew Blumberg offered a measured assessment. The results are "pretty short" constructions — "either counterexamples or small, extremely clever constructions that build on known things" — rather than deep conceptual breakthroughs. "If a person had done many of these things, that person would be justly lauded," he told Mashable, "but they don't change my priors."mashable