Summary:
En juillet 2025, la Maison Blanche a publié le Plan d’Action sur l’IA, marquant un tournant dans la politique américaine en matière d’IA. Ce plan, qui se concentre sur la domination technologique mondiale, comprend trois piliers : accélérer l’innovation, construire l’infrastructure IA américaine, et diriger la diplomatie internationale en matière d’IA. Accompagné de trois décrets exécutifs, ce plan vise à renforcer la sécurité nationale et la compétitivité économique des États-Unis.
Original Link:
Original Article:
Just when you thought July was coming to a quiet and serene end
Maybe you were even planning your pistachio gelato and Aperol Spritz beachside combo…
Then, the White House drops America’s AI Action Plan.
The AI Action Plan represents a fundamental shift and reshaping of U.S. AI policy.
The core focus is on “achieving and maintaining global technological dominance” in AI, as an economic and national security imperative.
The Biden administration’s focus on AI governance and safety, as well as tightening AI export controls, is firmly left in the past.
The AI Action Plan is a comprehensive framework consisting of over 90 AI policy recommendations, across 3 core pillars:
1. Accelerating Innovation
2. Building American AI Infrastructure
3. Leading in International AI Diplomacy and Security
The AI Action Plan is not law. It outlines the policy priorities of the White House. However, it was accompanied by 3 new Executive Orders, that are legally binding instruments:
1. Accelerating Federal Permitting of Data Center Infrastructure
2. Preventing Woke AI in the Federal Government
3. Promoting the Export of the American AI Technology Stack
Back in January, Trump revoked Biden’s Executive Order on AI safety. Then, a new Executive Order was issued, EO 14179: Removing Barriers to American Leadership in AI.
The stated policy of the current administration is to “sustain and enhance America’s global AI dominance in order to promote human flourishing, economic competitiveness, and national security”.
EO 14179 mandated the formulation of an ‘AI Action Plan’ by July 2025. It was published this week.
In next week’s newsletter post, I will do a deep-dive analysis on the AI Action Plan and the 3 new Executive Orders.
It’s been a whirlwind month for AI governance.
Here is an updated recap of everything that has happened in AI governance in July 2025 (and we still have 6 days left):
➡️ Senate votes to reject 10-year moratorium on U.S. state AI laws
➡️ BRICS sign declaration on AI governance, urging for UN-led frameworks
➡️ EU announces there will be no EU AI Act delay
➡️ EU publishes General-Purpose AI (GPAI) Code of Practice
➡️ OpenAI and Anthropic announce they will sign the GPAI Code of Practice
➡️ Meta announces it will not sign the GPAI Code of Practice
➡️ EU AI Office publishes GPAI Guidelines
➡️ New Zealand publishes National AI Strategy
➡️ UK government minister announces upcoming consultation on UK AI legislation
➡️ UK government minister announces upcoming consultation on UK AI legislation
➡️ EU publishes training data transparency template for GPAI model providers
➡️ White House publishes AI Action Plan