Countries / United States

AI in the United States

Build fast, regulate lightly at federal level, and push one national rulebook that overrides state AI laws.

Profile reviewed 24 Sep 2026323 developments tracked

Stargate AI infrastructure pledge (Jan 2025)
$500bn
California’s frontier-model law, SB 53, in force
1 Jan 2026
Genesis Mission federal commitments (Jul 2026)
$5bn+

Updated daily · newest first

Latest from the US

Adoption Anthropic will relax Claude’s biology safeguards for verified labs Verified life science teams will get access to Claude models with looser limits on biology work, monitored against a declared use case rather than blocked in real time. Applications · Policy Policy Anthropic puts paid outside evaluators inside its own building Anthropic will give evaluators from Accenture employee-level access to its model development, with no published standard yet for what they may report. Policy Research Anthropic used Claude to speed up 30 open biology models Anthropic reports roughly 4x average speed-ups after Claude spent four weeks optimising more than 30 open-source biology models — all figures its own. Research Invention OpenAI launches Astra for Law, with its own benchmark numbers OpenAI’s new legal configuration of GPT-6 Astra passed 54.0% of a research benchmark it ran itself, against 38.7% for the same model with plain web search. Applications · Models Research Anthropic puts a number on how much AI research Claude leads Anthropic says Claude now leads 26% of its own AI research work, up from under 1% in February, and wants rival labs to publish comparable figures. Policy · Research Research Google open-sources a way to fake the freight data nobody will share MilleMiglia generates synthetic freight networks so researchers can benchmark middle-mile logistics without the proprietary data carriers keep private. Research Adoption NVIDIA puts megawatt figures behind running AI clusters below full power NVIDIA has published operating figures for capping GPU power across a cluster, including a megawatt shed in under a minute — vendor numbers, but unusually specific ones. Infrastructure Research A proof about what shallow transformers cannot compute An IBM Research team has proved that constant-depth transformers and diffusion language models cannot match shallow quantum circuits on two specific tasks. Research

Profile

Reviewed monthly · Sources (6)

Strategy

America’s AI Action Plan (July 2025) sets out 90+ federal actions on innovation, infrastructure and diplomacy. The Genesis Mission for AI in science has $5bn+ in federal commitments across 278 projects (July 2026).

Regulation

No federal AI law. Executive Order 14365 (Dec 2025) set up a DOJ task force to challenge state AI laws. California’s SB 53 on frontier-model transparency has applied since 1 Jan 2026; Colorado replaced its AI Act with a narrower law from 1 Jan 2027.

Compute

Stargate targets 9+ GW across seven US sites by 2029, with Abilene, Texas the first live site (Epoch AI). Nvidia H200 exports to China have been licensed case by case since January 2026.

Key labs

OpenAI, Anthropic, Google DeepMind, Meta, xAI, Microsoft, Nvidia.

Notable models

Every frontier release is tracked on Model releases; the September 2026 flagships are GPT-6, Claude Opus 5.5, Gemini 3.8 and Grok 4.7.

What to watch

The first DOJ lawsuit against a state AI law; SB 53’s first year of enforcement; whether the Abilene site reaches 1.2 GW in Q4 2026.
Sources
  1. America’s AI Action Plan (White House)
  2. Genesis Mission update (White House)
  3. Executive Order 14365 (Paul Hastings)
  4. Colorado repeals and replaces its AI Act (Skadden)
  5. California SB 53 explained
  6. Stargate: where the US sites stand (Epoch AI)

Who builds AI in the US

All companies →

the US on the timeline

Full timeline →
  1. Waymo launches in Denver, San Diego and Tampa
  2. The DOJ backs OpenAI in the New York Times case
  3. OpenAI opens its agent harness as a managed API
  4. Anthropic names the misuse operations it disrupted
  5. OpenAI puts conversational ads inside ChatGPT
  6. Anthropic lets vetted labs past its biology safeguards