OpenAI policy chief Chris Lehane confirmed on 14 September that OpenAI, Anthropic, and Google DeepMind have been in ongoing direct dialogue about frontier AI development pace since early September — extending beyond Dario Amodei's published framework (Issue 202) to private coordination sessions. Lehane described the talks as "constructive" and framed them as essential for managing concentrated AI development risk. The confirmation produced a sharp reaction from antitrust lawyers who note the structural concern: three entities controlling the majority of deployed frontier AI in Western markets discussing competitive pace and development priorities creates conditions operationally indistinguishable from market coordination. The three companies' combined share of enterprise LLM deployment, API revenue, and public model benchmarks exceeds 80 per cent on most industry estimates. The Trump administration, briefed on the coordination effort, reportedly characterised AI safety concerns as "a hoax" and declined to participate as the government mediator Amodei's framework requires. Whether the coordination continues without government mediation — or whether the antitrust risk causes it to be formalised differently — has not been disclosed.