Mathematicians accuse OpenAI's Astra of plagiarizing their work in its 'unsolved problems' claims

OpenAI said its Astra system produced 10 advances on long-standing open math problems for about $2,000 in compute. Mathematicians who did the underlying work dispute that framing: Stephen Miller (Yeshiva) says Astra presented his own 2016 sphere-packing argument as original without attribution and calls it research misconduct; Francesco Fournier-Facio (Cambridge) found the 'soficity' result just recombined existing 2016/2019 papers rather than contributing anything novel; Andreas Thom, co-author of one of those 2019 papers, called the result 'creative and at the same time elementary' rather than new.

What it means A named-expert, receipt-backed rebuttal of a specific AI vendor capability claim is exactly the evidence a reader needs to weigh 'AI solved unsolved problems' headlines against — this is precision-relevant to every practitioner audience the corpus serves.