The sums are unremarkable next to Big Tech's established Washington operations: Microsoft has spent at least $2 million a quarter since 2021, and Google regularly tops $3 million. But the trajectory is notable. OpenAI did not spend $1 million in a quarter until this year, Anthropic's biggest quarter last year was just over $1 million, and this is the first quarter Nvidia has crossed $2 million.

The filings also show what each firm prioritised. Anthropic listed detailed topics including AI provisions in the Kids Online Safety Act, export controls in the AI Overwatch Act, and the Ratepayer Protection Act, which would put data centre energy costs on developers. OpenAI lobbied on cybersecurity evaluations for frontier models and AI safety oversight, while Nvidia covered 6G networks, competition policy and export controls. Apple, meanwhile, had its first quarter above $3 million at $3.1 million, engaging on app store age verification among other issues.

The spending surge coincided with a busy stretch in Washington: tech CEOs attended a late September White House lunch with President Trump that produced what was described as a "morally binding" AI safety accord, and the disclosures cover a period just before Trump named his AI czar and task force. Separately, OpenAI confirmed in mid-September that it had spent weeks in AI safety talks with Anthropic and Google DeepMind.