Archive BRAID DAILY
The White House governs AI one company at a time
Subscribe

Braid Daily · 2026-06-18

The White House governs AI one company at a time

Axios reports a case-by-case operating model behind this month's access fights.

Dark editorial illustration of a government table with a row of levers, one lit in signal yellow, evoking case-by-case AI policy control.

The lead

1

Following this month's Anthropic export-control story, Axios reports the through-line. Voluntary standards get announced in public. The real decisions — who can run which models — get made case by case. Two Axios pieces today put Lutnick and a 'shadow' executive-action track at the center of how access gets decided.

Read source
Diagram contrasting public voluntary AI standards with case-by-case White House pressure, and a proposed mandatory-vetting bill.
The reported operating model: voluntary standards announced in public, access and export decisions made company by company.

The policy operating model

4

Trump's shadow AI policy

Axios

A companion piece on governing through executive action rather than legislation. Read it alongside the power-center story to separate what's enacted from what's reported pressure.

Read source

Where the machines can actually run

3

The access layer outside Washington

4

On the agent-reliability shelf

4

Talent flows

1

Companion episode

Who Gets to Approve the Model

· 00:22:51

We've tracked the Anthropic cutoff as an export-control story all week. Today's Axios reporting puts it inside a broader pattern: standards announced in public, access decided in private. Worth watching whether Gottheimer's mandatory-vetting bill turns any of that into law.