OpenAI’s “AI in America” blueprint is really a list of demands for the US government
Framed as a plan for ensuring American superiority in AI, the document warns that overregulation will drive hundreds of billions of dollars to Chinese AI projects.
OpenAI just published a 15-page manifesto titled “AI in America: Open AI’s Economic Blueprint.” But if you read between the lines, the blueprint boils down to a wish list of things that OpenAI wants from the US government:
🚦 Voluntary “rules of the road” instead of federal regulation
⚖️ Exclusion from the patchwork of state AI regulations
🪖 Classified national-security briefings
🚔 Defense, national-security, and law-enforcement contracts
🎟️ Exclusion from any AI regulations if the companies work on national-security applications
📊 Mountains of digitized government data to train its AI on
🍎 Public-school technology-budget dollars
🏛️ State-government-agency contracts
🎓 State-university research dollars
🧪 Federal science-research dollars
©️ Freedom from copyright restrictions
☢️ Fast-track permitting process for nuclear reactors and other energy generation
⚡️ Energy updates and energy infrastructure for powering data centers (including fusion, championed by Sam Altman’s startup, Helion)
🏭 Domestic chip manufacturing
🇺🇸 Federally backed AI-company capital expenditures
And finally, OpenAI says if doesn’t get these things, America just handed AI superiority to the Chinese Communist Party.
Framed as a plan for ensuring American superiority in AI development, the document warns that moving too slowly or regulating the industry too much will drive hundreds of billions of investment dollars to Chinese AI projects. The cautionary tale told by OpenAI is how the UK ceded the automobile boom to the US by overregulating, despite introducing some of the earliest cars.
In the past, OpenAI actually asked the US government to create some regulations for AI, but this document calls for a lighter touch, calling for “common-sense rules of the road,” voluntary AI risk-assessment programs, and “best practices” in place of regulation. In fact, it calls for creating a loophole for AI companies to avoid regulation entirely, if they agree to work on national security with the government (which OpenAI, Meta, and Palantir are already conveniently doing).
It even calls for the government to give classified national-security briefings to AI companies like OpenAI. OpenAI also calls for export controls for frontier AI models to allow use by partners and allies of the US, while limiting access to “adversary nations.”
The document highlights the need to ensure AI companies protect children by “encouraging” ways to prevent child-sexual-abuse material from being created or distributed, along with working closer with law-enforcement agencies. OpenAI also calls for audio and video generated by AI to include “provenance data” to build trust.
Much of this blueprint outlines a sprawling proposal for billions of federal dollars from every corner of federal and state government. From the American public-school system, the document calls for “robust funding for pilot programs, school district technology budgets and professional development trainings that help people understand how to choose their own preferences to personalize their tools.”
OpenAI also sees a new market for selling its services to state governments by “supporting experimentation with AI, including by start-ups and smaller AI firms, to identify ways to solve people’s daily hard problems in areas like education and healthcare.”
Infrastructure destiny
In a section titled “Infrastructure as Destiny,” OpenAI opines that the US government has no choice but to help AI companies by building massive amounts of energy infrastructure as well as domestic chip manufacturing. The document suggests that the government should invest in next-generation energy technology, including fusion, which is being pushed by OpenAI CEO Sam Altman’s energy startup, Helion.
Noting the “massive amount” of capital needed for building out the AI infrastructure to ensure America’s domination of the field (Altman has previously floated a $7 trillion figure), OpenAI is calling for “federal backstops for high-value AI public works,” which sounds like a government guarantee of spending that would include classifying AI data centers as “national strategic assets.”
The document says:
“In the AI era, chips, data, energy and talent are the resources that will underpin continued US leadership, and as with the mass production of the automobile, marshalling these resources will create widespread economic opportunity and reinforce our global competitiveness.”
