Maria Cury, tech policy reporter at Axios, says the Sanders and Bannon partnership is a sign of how artificial intelligence has scrambled the political map and political norms.
Page F5From§Eachthe supper edition — 14 September 2026
White House quietly drafts AI extinction safeguards days after Trump dismissed the same warnings as a hoax
As it ran on the front
Reconciliation, Monday, September 14. The entry in the public ledger: the President called the extinction warnings a hoax. The entry in the private ledger, filed the same day by a White House source to the New York Post: advisers are scrambling to draft "limited safeguards" against AI apocalypse and mass extinction. Same date, same building, two ledgers.
The warning he called a hoax came from the chief executives of Anthropic and OpenAI, the companies building the technology his own advisers are now moving to regulate. Vice President JD Vance filed his own entry that day, too, calling the same warnings a "Trojan horse." A horse, a hoax, and a safeguard, all drafted before the news cycle turned over.…
…(cont) Credit where it's due: the debate did not start Monday. By September 11, three days earlier, a former Anthropic researcher's warning had already pushed past the trade press and into general circulation — "last week," The Hill's sourcing put it, meaning the industry had been arguing this out before the President had an opinion on it. The hoax call arrived late to a conversation already underway.
Congress logged its own split the same day. Senate Majority Leader John Thune wants a "light touch," his words, so as not to "stifle innovation," also his words — the standard filing whenever the ask is: please do less. House Minority Leader Hakeem Jeffries wants the opposite: cancel recess, stay in Washington, take up limits on AI companies before anyone leaves for the districts. One side wants speed preserved. One side wants the calendar cleared to slow it down. Both are citing the same warnings.
The stranger filing that day: Senator Bernie Sanders and Steve Bannon, standing together for human-controlled AI, a pairing Axios's tech policy desk read as a sign the technology has scrambled the political map. Note what the ledger does say and what it does not. It says the map moved. It does not say why, and neither does this column — that entry stays open.
Last item, same day: Nvidia's Jensen Huang, onstage in Los Angeles, got a phone call from the President, who used it to criticize calls to slow down and regulate the very industry Huang runs. The hoax, the safeguard, the light touch, the recess fight, and the phone call all landed on September 14. Print the whole file and it runs thirty pages. The gap between the podium and the memo, hon, is one sentence.
“He calls it a hoax on camera, and by the same afternoon his own people are drafting safeguards for the extinction he just said isn't real. That's not indecision — that's two different answers for two different audiences, the cameras get the hoax and the building gets the memo. Figure out which one you're standing in.”
“Look, 'limited safeguards' is what a light touch looks like — Thune's right, you don't stifle innovation, you just also quietly draft rules against mass extinction while calling the whole warning fake on television. That's not a contradiction, that's balance. Did I just call calling something a hoax "balance"? Let me circle back on that one.”
The receipts
President Trump flatly claimed it was a “hoax” on Monday after the CEOs of Anthropic and OpenAI said artificial intelligence risks destroying the world – but White House advisers are nevertheless scrambling to devise a balanced policy to regulate AI, On The Money has learned.
Mr. Trump called Jensen Huang, Nvidia’s chief executive, while Mr. Huang was onstage in Los Angeles and criticized calls to slow down and regulate A.I.
Vice President JD Vance called the warnings from AI leaders a "Trojan horse."
House Minority Leader Hakeem Jeffries (D-N.Y.) is calling on House GOP leaders to remain in Washington to allow Congress to consider new limits on AI companies following dire warnings from industry insiders about the threat the technology poses to humanity. Speaking to reporters Monday in the Capitol, Jeffries said the calls for a…
Senate Majority Leader John Thune (R-S.D.) on Monday said any legislation Congress passes to regulate AI must have a “light touch” so as not to stifle innovation and tempered recent warnings that superintelligent AI could exterminate humanity as a “worst-case scenario.” “It needs to strike a balance, light touch.
· The debate over AI 'doomsday' warnings from the morgue, 11 Sep 2026
This page is a back-issue: the story as it ran, receipts as they were. The current edition is at the front. The byline is a pen name for a column drafted by a machine and checked by the editor: how this is made.