• Sources: primary, repost, discussion
  • Summary: The association's own statements page carries its statement on OpenAI's 2026-10-06 release of mathematical documents, and Terence Tao's blog carries the same text as a guest post. The statement says the release ignored the Advisory Group on Mathematics and Artificial Intelligence position that frontier labs should not test advanced mathematical problems on internal models, and it asks mathematicians to stop working with OpenAI.
  • Why it matters: Expert review of model output on advanced mathematics becomes something labs may have to source without the cooperation of the field.