Trump's answer to the pacing ask: he already has criminal power over the labs

Share
Trump's answer to the pacing ask: he already has criminal power over the labs

The AI pacing debate stopped being about risk on Monday. Washington's reply to the frontier labs asking for a rulebook was a reminder about the powers already on the books, and from Frankfurt came the counter-argument that Europe is buying the leverage it claims to lack.

President Trump spent Monday telling the AI industry it needs no new guardrails, then listing the ones he says he already holds. In an all-caps Truth Social post, he wrote that "the only control or 'guardrails' that AI needs is a STRONG AND SMART (High IQ!) PRESIDENT," that "We already have tremendous CRIMINAL and REGULATORY power over these companies," and that "There is a SICK conspiracy going on against AI and Data Centers, and the only one that is happy about it is China." He named Anthropic's Dario Amodei directly, claiming the administration "has stopped AI 'people' from doing bad, or potentially bad, 'things,'" and that Amodei "is now pretending to be a 'perfect little angel.'" He did not name the conspirators, describe the criminal authority, or say what conduct he meant; the post closed with "Conspiracy Theorists, Treasonists, Traitors, and Leakers, BEWARE!" Vice President JD Vance gave the same idea a softer shape for reporters, saying it feels "a little bit weird" that so many frontier companies are "coming to the government and begging the government to regulate them... like a little bit of a Trojan Horse."

The read: this is not a light-touch posture, it is an enforcement posture with no statute behind it. The labs asked for rules and got a DOJ reminder — which bites precisely because a coordinated slowdown is the arrangement antitrust law exists to prohibit. That is why OpenAI asked Congress whether industry-wide pacing would even be legal, and why David Sacks, who now chairs the president's science advisers' council, has told both Anthropic and OpenAI they are free to pace themselves but should stop demanding an antitrust waiver as the price. A slowdown without legal cover can only happen one company at a time, which is the exact coordination problem Amodei's essay proposed solving. Markets kept score regardless: Nasdaq futures fell more than 1% overnight, Intel dropped close to 6%, Micron about 5% and Nvidia roughly 3%. We tracked the company response this morning — Altman puts OpenAI's pacing on paper — and drops the legal caveats.


Europe's most senior financial official reframed AI dependency Monday as a balance-sheet problem rather than an industrial-policy wish list. Speaking in Vienna, ECB president Christine Lagarde said Europe must become a producer of AI rather than an importer, because a technology soon "screening goods at the border, deciding which tax returns are audited, dispatching trains, watching patients on wards and clearing payments at banks" is a dependency that can be switched off all at once. Her framing was about bargaining power: "A withdrawal of access, or a change in its terms, would then reach every sector at once... That is leverage of a kind no trade partner has ever held over Europe, and it could be used in any negotiation, on tariffs or on digital taxes." The diagnostic numbers were blunter than most EU AI speeches: Europe already has too little data-centre capacity to meet its own demand and, on current trends, that gap is projected to grow more than sixfold within a decade. Adapted quickly, she said, AI could lift productivity by up to 4% over ten years — transformative for public finances — so the goal is not frontier models but ones "good enough" for most tasks, running on European infrastructure, until the threat of being cut off stops working.

The new part is the money channel. Lagarde said US technology firms' financing needs are so large that they are borrowing in Europe, crowding other issuers out of the debt market and pushing up everyone else's costs, and that European pension funds' heavy positions in US tech stocks mean any correction lands directly on European savings. That is a financial-stability case for sovereign compute, delivered by the person who sets eurozone rates, one week after the Draghi line that Europe holds roughly 5% of global AI compute — Draghi tells Europe to build AI data centres — it holds 5% of compute — and against a backdrop of tariffs, a renewed US push for Greenland and American troop withdrawals from Europe.

What to watch: whether any eurozone institution turns the crowding-out claim into an actual financing vehicle, and whether the September 24 Trump-Xi meeting produces a governance line that makes either speech obsolete.

If the choice is between a slowdown the labs cannot legally coordinate and a buildout Europe cannot finance, whose brake pedal are you actually betting on?

Read more

OpenAI drops 372 math results, nearly all from one prompt

OpenAI drops 372 math results, nearly all from one prompt

A dense news day across three fronts: OpenAI published a mountain of mathematics from an unreleased model, SpaceX reportedly wants the debt market to fund its Nvidia orders, and Anthropic restructured its cyber program around three tiers of trust. OpenAI published 372 mathematical result families from an internal frontier model — and says nearly all of them came from a single prompt handed to a single AI agent. The release comprises 722 manuscripts touching hundreds of open questions, plus Lea

Underdog launches a private on-device AI assistant, backed by a16z

Underdog launches a private on-device AI assistant, backed by a16z

The privacy split in consumer AI got a new entrant tonight, GitHub turned code-review benchmarks into a vendor-bias argument, and OpenAI opened its oddest API to everyone. Underdog launched in invite-only beta: an on-device AI assistant that keeps your data on your machine and never charges you a subscription. Self-taught coder and Thiel Fellow Sigil Wen — who moved to Silicon Valley at 17 and lived in an AI hacker house with Andrej Karpathy — built his own inference engine, Husky, to run a 2

Today in AI — October 6, 2026

Today in AI — October 6, 2026

A day of second-order moves: the labs are buying task data from software vendors instead of scraping the web, Waymo is borrowing to fund the robot world, and the biggest bank in the US just put a number on what Anthropic's latest model cost it in risk. Models & Research * OpenAI is training GPT-6 Astra on Ironclad's real contracting work. Ironclad staff helped turn 11 tasks across legal, commercial and procurement work — setting up NDAs, approval workflows, clauses that change by jurisdict

Meta, Walmart and Stripe publish the Personal Agent Protocol

Meta, Walmart and Stripe publish the Personal Agent Protocol

The agent economy is writing its rulebook tonight: one open standard for AI bots at the checkout, a cheaper image model from Google, and Anthropic turning bug-hunting into a tiered product. Meta, Walmart, Stripe and Sierra are publishing an open "personal agent protocol" — a standard that defines how personal AI agents interact with businesses online. The group behind it reads like a cross-section of agentic commerce: Meta, Sierra, Walmart, Stripe, Shopify, Genesys, Rocket, NiCE, Decagon and I