The House adjourned Wednesday and fled Washington to campaign, punting AI regulation past the midterms — and that may be the best thing that happened to American innovation all year.

The loudest voices demanding Congress "do something" on AI are the very corporations that stand to lock in their dominance. What insiders call "guardrails" looks more like a moat: government-mandated auditors, Commerce Department power to suspend models, and an antitrust exemption so the biggest AI firms can openly collude. The real scandal isn't that Congress left town. It's what they'd pass if they stayed.

The urgency campaign started when an Anthropic researcher resigned and claimed the technology "could kill us all by the end of the decade," according to CNBC. Anthropic, OpenAI, and xAI have practically begged Washington to regulate them. When the companies demanding oversight are the ones with the lawyers and lobbyists to shape it, that's not safety — that's regulatory capture.

House Speaker Mike Johnson canceled Thursday's planned work and sent members home. "It's time for our members to go home to their districts and make their case to the American people," Johnson said. He also warned that rushing to slow AI progress could hand China a boon in the race for AI dominance. President Trump has called reported AI threats a "hoax" and opposed restrictive regulations.

Not everyone is pleased. Rep. Sam Liccardo, D-Calif., led 107 Democrats in writing to Johnson demanding the House stay. "While AI safety experts and Americans increasingly urge action to confront this conflagration of risk, Congress fiddles," the letter read. Liccardo complained the House wasted floor time on "performative resolutions opposing socialism" before taking another six-week vacation.

Liccardo's "low-hanging fruit" tells you everything about where this goes. He wants the Frontier Act, a bipartisan bill from Reps. Jay Obernolte, R-Calif., and Lori Trahan, D-Mass., that would require third-party auditors for AI labs and give the Commerce Department power to suspend or restrict a developer's model posing "imminent catastrophic risk." He also wants a "kill switch" provision to shut down AI models and a narrow antitrust exemption letting top AI companies collaborate on safety — which is to say, letting incumbents coordinate without competitive consequence.

Obernolte himself, along with Rep. Brian Fitzpatrick, R-Pa., joined Democrats in urging leadership to advance guardrails "as soon as practicable." Obernolte told reporters: "The urgency of the moment requires action by Congress before the end of the year."

Here is the question neither party's regulators want to answer: if AI truly poses an existential threat, why is the first instinct to give the Commerce Department — an executive-branch agency — unilateral power to pull the plug on any model it deems dangerous? And if it doesn't, why are we restructuring an entire industry around the panic of the companies already on top?

The Senate remains in session for two more weeks. Whether anything moves there — or in a post-election lame duck — is the open question.