Coxon's Anthropic exit triggers Senate probe of OpenAI

Coxon's resignation did not come out of nowhere. It landed after a decade in which nearly every federal AI safety effort stalled. The FUTURE of AI Act was introduced in 2017, before Congress even tracked "artificial intelligence" as its own category, and went nowhere. Rep. Yvette Clarke's DEEP FAKES Accountability Act followed in 2019 and also stalled. The one thing that did pass in 2020 was folded into that year's defense bill to fund research and workforce training — competitiveness spending, not regulation.

The pattern held even as the stakes rose. Sam Altman testified in May 2023 asking to be regulated, and produced zero legislative text. Chuck Schumer's nine closed-door AI Insight Forums drew 108 attendees, 44 of them from industry, and yielded a roadmap but no bill. California's SB 1047 passed the legislature in August 2024 only to be vetoed that September. Biden's Executive Order 14110 survived 14 months before Trump revoked it on his first day. Meanwhile the industry's political playbook hardened: more than $40 million in outside money flooded Alex Bores's June 2026 primary, and Colorado's most ambitious AI law was gutted five weeks after xAI sued.

What Coxon's resignation appears to have changed is the political cost of silence. Within 48 hours, Hawley opened a Senate investigation, Blumenthal sent a nearly identical letter, Sanders convened a briefing with Hinton, and Khanna proposed more legislation. The test now is whether the Cruz-Klobuchar-Thune bill's safety provisions survive Democratic demands, and whether its likely preemption of state AI laws becomes the price of federal action. For the labs weighing highly anticipated IPOs, the outcome hinges on whether they can shape the text before it leaves committee.