Quote · Pivot
The White House UFC Fight, SpaceX’s Big Pop, and Fox’s Roku Deal
Where this was said
Anthropic's 90-Minute Shutdown: National Security or Political Hit?
At 46:02 · chapter starts 43:40
On a Friday, the Trump administration issued an order giving Anthropic just 90 minutes to take down its flagship models, citing a claim — apparently originating from Amazon's Andy Jassy — that a researcher used the model to assist in cyberattacks. [1] — Kara Swisher "The Trump administration ordered Anthropic to take down its most powerful AI models in 90 minutes, citing a national security threat. Kara …" 43:40 Anthropic dispatched staff to Washington to dispute the action, arguing correctly that the same jailbreaks exist across all major AI models, not just theirs. Kara and Scott are skeptical of the official rationale: Dario Amodei has been a persistent thorn in the administration's side, refusing to go along with requests around weapons development and privacy violations, and both hosts read the shutdown as a political hit coordinated with pro-administration Silicon Valley interests. Scott relays that Ian Bremmer told him the Anthropic shutdown was a bigger topic at the G7 than the Iran memo — because foreign companies and NATO defense departments that had deeply integrated Anthropic into their workflows suddenly faced an arbitrary kill switch. Scott simultaneously applauds crisp government action on AI safety and says he doesn't trust these particular people's motives. Kara calls David Sacks' justification pompous and unconvincing. The hosts also note the separate but related wave of lawsuits hitting AI companies, including a Florida suit against OpenAI and a multi-state AG investigation, as the more structurally serious legal threat.
The Trump administration ordered Anthropic to take down its most powerful AI models in 90 minutes, citing a national security threat. Kara Swisher and Scott Galloway suspect it's really a political kneecapping of a company that refused to play ball with the administration — and the chaos it created for foreign companies and NATO allies is very real.
Scott Galloway calls for a bipartisan panel of economists, philosophers, and tech experts to screen all AI model releases through a 30–90 day review — identical for every company. Regulatory certainty is good for business. What's not good is one political donor's company getting a kill switch activated while competitors are left alone.
