Karp: AI Giants Want Rules So They Can Point to Them Later

Palantir CEO Alex Karp says the push by major AI labs for federal regulation isn't about locking out smaller rivals. It's about swapping legal exposure for a rulebook. Karp's remarks, replayed and debated on a recent podcast panel, argue that AI companies are trying to escape what he calls the "first line of defense": liability for their own actions.

Karp's framing, according to the clip discussed on the show, works like this. Right now, a company that builds a technology capable of catastrophic harm carries civil and criminal liability by default. Disclosure comes first — what did you know, and what did you do about it. Regulation, in his telling, is the industry's attempt to substitute a government checklist for that open-ended exposure: follow the rules regulators write, and use compliance as a shield if something goes wrong.

Karp also flagged a staffing problem for any regulatory regime: the people qualified to write and enforce AI rules mostly work for AI companies. He contrasted that with the scientists who built nuclear weapons, who he says acted believing they weren't on anyone's payroll.

The panel picked up Karp's thread and ran further than his transcript supports, floating a dollar figure — one host suggested AI firms fear being "on the hook for like a trillion dollars of damage" — and drawing a comparison to COVID-origin accountability. Those are the hosts' extrapolations, not claims Karp made in the aired excerpt, and Boardroom is not adopting them here.

Karp did not name specific companies or specific regulatory proposals in the portion aired. The episode's exact broadcast date and number were not specified in the material reviewed for this brief.

Standing line: fewer rules, clearer numbers, no cheering section for founders.

Heard on The Armstrong & Getty Show — Nay Means Nay (2026-09-18).