Trump's AI Action Plan is a blueprint for dystopia

The president is trying to give Silicon Valley nearly everything it wants — at a price.

Alright, I’ve officially spent too much time reading Trump’s 28-page AI Action Plan, his three new AI executive orders, listening to his speech on the subject, and reading coverage of the event. I’ll put it bluntly: The vibes are bad. Worse than I expected, somehow.

Broadly speaking, the plan is that the Trump administration will help Silicon Valley put the pedal down on AI, delivering customers, data centers and power, as long as it operates in accordance with Trump’s ideological frameworks; i.e., as long as the AI is anti-woke.

More specifically, the plan aims to further deregulate the tech industry, penalize US states that pass AI laws, speed adoption of AI in the federal government and beyond, fast-track data center development, fast-track nuclear and fossil fuel power to run them, move to limit China’s influence in AI, and restrict speech in AI and the frameworks governing them by making terms like diversity, inclusion, misinformation, and climate change forbidden. There’s also a section on American workers that’s presented as protecting them from AI, but in reality seeks to give employers more power over them. It all portends a much darker future than I thought we’d see in this thing.

In his speech—though not in the plan itself—Trump advocated for excusing AI companies from copyright laws, so as to better compete with China, in comments that echo a key Silicon Valley talking point.

Before we dive into the nuts and bolts of what’s being proposed here I want to pause a consider a few questions. Like: What does it mean that this AI plan was unveiled just a week after the Trump administration dissolved half of the Department of Education? Why is it that, after gutting the nation’s funding for science across the board, and expressing hostility to cleantech like solar panels and electric vehicles, is Trump willing to earmark funds for AI research and development? Why is a widely anti-tech Trump administration so enthusiastic about this technology in particular?

One reason is, sure, that his administration has close ties to Silicon Valley; Elon Musk, Marc Andreessen, Peter Thiel, Palmer Luckey, and David Sacks and their stores of cash helped deliver him the election. Sacks, a bigwig venture capitalist, is now Trump’s AI czar, and one of the authors of the AI action plan. Another author is Michael J. Kratsios, who was the managing director of Scale AI (recently absorbed by Meta), right up until he took the job as chief technology officer to the president.

But it’s even more than that. As JD Vance was early to recognize, AI dominance promises power. It promises technology that can surveil, automate, bring workers to heel, and create cheap and voluminous propaganda. Anyone and everyone who’s bullish on AI should ask themselves why it is that the AI industry, in its current formation, is so attractive to, willing to partner with, and such a natural fit for a regime that grows more authoritarian by the day.

Finally, it’s worth noting that despite the near-universal coalescence around AI as The Future in Silicon Valley, AI companies are burning through cash at a rather unsustainable rate. If current trajectories hold, companies like OpenAI and Anthropic may have to take drastic measures—like, for instance, turning to increasingly expansive government contracts of the sort outlined in the AI Action Plan. In its current state, Silicon Valley likely needs Trump as much as Trump needs it.

Here’s the part where I issue a brief reminder that BLOOD IN THE MACHINE is a fully independent publication, and my work is made possible 100% by paid supporters. If you find value in this work, and you’re able to, consider becoming a paid subscriber. You’ll get access to what I’ve been calling the Critical AI Report, a roundup and analysis of key AI stories, as well as my near-infinite gratitude. Thanks, and onwards.

Edited by Mike Pearl.

So, first—here’s the gist of what happened yesterday, as reported by the New York Times:

Mr. Trump signed three executive orders and outlined an “A.I. Action Plan,” with measures to “remove red tape and onerous regulation” as well as to make it easier for companies to build infrastructure to power A.I.

One executive order barred the federal government from buying A.I. tools it considered ideologically biased. Another order would speed up the permitting process for major A.I. infrastructure projects, and a third focused on promoting the export of American A.I. products around the world.

“America is the country that started the A.I. race,” Mr. Trump said in a Wednesday evening speech in front of administration officials and tech executives, including Jensen Huang, the chief executive of the chipmaker Nvidia. “And as president of the United States, I’m here today to declare that America is going to win it.”

Because I can’t help but notice these things, here is the soundtrack selected for the minutes leading up to that Wednesday evening speech, in the order that they were played:

• Thriller, by Michael Jackson

• Whole Lotta Shaking Going On, by Jerry Lee Lewis

• Burning Love, by Elvis

• Johnny B Goode, by Chuck Berry

• The Christian hymnal How Great Thou Art

• Nothing Compares 2 U, by Sinead O’Conner

Pretty weird playlist for an AI speech imo! You have to think that Sinead would have objected to being included there.

Trump came out to Lee Greenwood’s God Bless the USA, and kicked things off with his trademark banter: “What a group of smart ones we have in front of me today,” he said, gesturing to the tech titans and AI CEOs in the front row. “That’s about as good as the comes, up here, the brainpower. [sic] The greatest power of them all: the brainpower.”

The speech was long and meandering, filled with asides and detours, only one of which was particularly notable, since it departed from the action plan.

AI companies should be free to steamroll copyright:

Previously the Trump admin has said it will leave AI matters up to the courts; Trump made it clear that he is in favor of letting AI companies vacuum up copyrighted works at will.

“When a person reads a book or an article, you've gained great knowledge,” Trump said. “That does not mean that you're violating copyright laws or have to make deals with every content provider. And that's a big thing that you're working on right now, I know. But you just can't do it. China's not doing it.”

Most of the meat is in the Action Plan itself, so that’s what we’ll dig into from here on out.

Introducing the state AI law ban, round 2

A chief thrust of the action plan is that we need to cut “red tape” that currently constrains AI development—I’m not exactly sure where this red tape preventing AI development is hiding, but the Plan is intent on getting rid of it. Trump’s plan calls on AI companies and businesses to write to the administration to tell them where they want fewer rules and restrictions.

More pointedly, the action plan states that

AI is far too important to smother in bureaucracy at this early stage, whether at the state or Federal level. The Federal government should not allow AI-related Federal funding to be directed toward states with burdensome AI regulations that waste these funds…

This is a reflection, and reiteration, of the recent, Silicon Valley lobbying effort to include in the budget bill an amendment that would have amounted to a moratorium on states’ ability to pass laws governing AI. The word on the Hill is that Ted Cruz plans to pursue similar legislation, perhaps in the Fall, and this language in the action plan is a signal that the administration is on board.

Free speech provisions that ban key words

In a particularly Orwellian segment of the AI action plan (and the thrust of one of the three executive orders), the Trump admin proposes that, in the name of free speech, words that it finds disagreeable, like “climate change” and “diversity,” be banned from AI management frameworks. Furthermore, AI services considered for federal contracts should be assessed by the government for ideological biases—also to ensure free speech is protected.

From the plan:

• … through the National Institute of Standards and Technology (NIST), revise the NIST AI Risk Management Framework to eliminate references to misinformation, Diversity, Equity, and Inclusion, and climate change.

• Update Federal procurement guidelines to ensure that the government only contracts with frontier large language model (LLM) developers who ensure that their systems are objective and free from top-down ideological bias.

Clear the way for more data centers and for more of the power plants needed to power them

This one’s pretty self-explanatory:

Create Streamlined Permitting for Data Centers, Semiconductor Manufacturing Facilities, and Energy Infrastructure while Guaranteeing Security

…AI will require new infrastructure— factories to produce chips, data centers to run those chips, and new sources of energy to power it all. America’s environmental permitting system and other regulations make it almost impossible to build this infrastructure in the United States with the speed that is required. Additionally, this infrastructure must also not be built with any adversarial technology that could undermine U.S. AI dominance.

The plan outlines a number of ways that Trump wants to make it easier to build out data centers, faster, and for energy companies to wipe away regulatory or environmental hurdles to build out new gas, nuclear, and even coal plants.

The plan follows Trump and Pennsylvania senator Dave McCormick’s announcement of $70 billion worth of energy and AI investments in the state—for a new spate of data centers and the plants needed to power them.

Accelerating AI adoption at work

The plan suggests the deployment of “sandboxes” or “AI Centers of Excellence” to encourage private industries—it singles out healthcare—that might otherwise be slower to adopt AI, to get on board. (It also proposes investing in AI research more roundly).

It also has whole sections devoted to “Accelerate AI Adoption in Government” and “Drive Adoption of AI within the Department of Defense.” Following the work begun by DOGE, the plan wants to see more of that AI-first strategy adopted across the government, especially in the DoD—meaning fewer public servants and more dubious AI systems.

A pro-boss AI worker policy

Probably because this newsletter focuses so much on work and workers and how they’re affected by technology, the “Empower American Workers in the Age of AI” section stood out. It’s presented, as you can see, as pro-worker, and notes that AI will “demand a serious workforce response to help workers navigate” the coming transition.

Alas, however, that “serious” response amounts to lip service and corporate handouts. It proposes job retraining programs for workers who lose their jobs to AI, but that’s about it. The plan proposes “clarifying that many AI literacy and AI skill development programs may qualify as eligible educational assistance under Section 132 of the Internal Revenue Code,” thus enabling “employers to offer tax-free reimbursement for AI-related training.” Basically, tax write-offs for offering AI trainings; another direct benefit to the Silicon Valley firms offering such products.

In short, it’s somewhat notable that the Trump administration felt the need to include a section for workers at all—but it winds up serving as an assuaging message to the executive class that Trump is prepared to help further propagate AI tools in the workforce, and to help deal with any of the fallout, should the tools wind up being disruptive to labor. To that end, the section helps emphasize that very point, letting employers know that the Trump admin is on board with the vision that AI can and should replace jobs, and will at least help pay lip service to assisting impacted workers.

The rest

There’s more—one of the most surprising parts of the plan is its nominal endorsement of open source AI development, for instance, though its unclear what impact said endorsement will actually have. (Or what the Trump admin really means by “open source.”) There’s language about needing to limit China’s influence with regards to AI, accelerating AI product exports, and so on. But the above is a pretty good summation of most of the key planks, I hope.

A blueprint for dystopia

The Action Plan is just a blueprint; none of it is legally binding. (The executive orders are a different story.) But together, with the executive orders, very recent history, and the presidents’ own off-the-cuff remarks, it offers, along with its policy guidelines, a pretty stark portrait of the AI-dominated future as Trumpworld sees it. It’s a future marked by a further erosion of democracy, and of AI-flavored authoritarianism.

The Trump administration is willing to put resources behind AI education and AI literacy, while it strips funds for actual public education. It’s willing to invest in AI research, while it strips funding for almost all other kinds of scientific research. It’s willing to sponsor the creation of vast data center infrastructure to give rise to the future, but seeks to kill incentives for solar and wind and bankroll nuclear and gas, ensuring AI will be propped up by the energies of the past.

Trump vocally supports Silicon Valley’s push for deregulation, and is on board with the push to ban states from passing AI laws. In other words, the plan will help de-democratize AI. Meanwhile, it aims to push AI into the hands of as many executives and managers as possible, speeding the entry of AI into workplaces both private and public. And needless to say, Trump is a fan of the technology—to date, he’s published at least three high-profile AI-generated images and videos: One of his team arresting Obama, one a Studio Ghibli-style image of a migrant woman being arrested by ICE, and one of Gaza being paved over and turned into a luxury resort.

@7newsaustraliaDonald Trump has posted a fake AI-generated video of former US president Barack Obama being arrested in the Oval Office, as Trump tries to redirect focus away from the Jeffrey Epstein controversy. #trump #donaldtrump #obama #barackobama #jeffreyepstein #arrest #president #us #usa #unitedstates #7NEWS

Tiktok failed to load.

Enable 3rd party cookies or use another browser

Trump’s AI plan is a tech industry wishlist with every box ticked neatly down the line. It signals the further coherence of Silicon Valley and the Trump-led federal government into a cooperative, automation-bent surveillance state. One where tech companies are assisted by the government in the project of deploying their for-profit AI systems onto every feasible surface.

It’s above all a blueprint for consolidated control, a vision where worker power withers, citizens become subservient to AI; to its corporate masters and federal overseers, where fossil fuels are ushered again into prominence at the height of the climate crisis, and democracy is sidelined in favor of automation technologies. It is, in summary, pretty grim!

There are countermeasures brewing. A call for A People’s AI Action Plan, which I had a small role in drafting, was raised and signed by over 110 organizations, from Data & Society to DAIR to the National Nurses United to the Writers Guild East to AI Now to this very newsletter. People are organizing, preparing to fight; we all should.

---