Bubeck responds to claims OpenAI tried to dictate authorship in Anthropic math collaboration
jasondeanlee · x · 2026-09-09
Sebastien Bubeck (now Anthropic, ex-OpenAI) publicly addressed accusations that he tried to control authorship in a joint math proof effort. The disputed screenshot, he says, was him reaching out to Microsoft's Levent to coordinate releases, and he denies ever asking for Levent's removal from his own work.
He explains that in a call with Tristan he learned the other side had only solved the Euler problem, not Navier-Stokes; afterward they brainstormed paths forward, including Tristan leading a rewrite of OpenAI's Navier-Stokes proof — the context in which he remarked "it would be simpler if Levent was not an Anthropic employee."
Researcher nanjiangcs pushed back, saying Bubeck unilaterally tried to dictate who from the other side could stay on the paper, and hinted at having had a similar experience.
More from Companies & People
- David Sacks questions Anthropic's credibility as AI industry's 'boy who cried wolf' — JacquesThibs · 2026-10-25
- Free EECS 6322 deep learning course: visual path from transformers to diffusion models — CSProfKGD · 2026-10-25
- Professor releases 'BACKPROP' trailer for upcoming visual deep learning lecture — CSProfKGD · 2026-10-25
- Every's 'frontier team' solves the keep-up-with-AI dilemma; Opus cuts explainer videos — every · 2026-10-25
- OpenAI's Shyamal Anadkat recruiting founding members for computation-physics-chemistry venture — shyamalanadkat · 2026-10-25
- OpenAI DevDay Exchange Berlin: DevRel team hosts ambassadors ahead of the event — paw_lean · 2026-10-25