COLUMN: Dario Amodei Joins the White House — and We Should All Be Worried
The Physicist Who Wanted to Save Humanity
Amodei isn’t Elon Musk. He doesn’t tweet at 3 a.m. He doesn’t sell cars. He has a Ph.D. in biophysics from Princeton, worked at Google Brain, and then at OpenAI as vice president of research. In 2021, he walked out along with his sister Daniela. The reason given: OpenAI was no longer taking safety seriously.
He founded Anthropic with a stated mission: to build an AI that is “honest, helpful, and harmless.” Three adjectives that, in 2025, sound more like a prayer than a business plan.
The Prophet of Doom
In October 2024, Amodei published a 14,000-word essay titled “Machines of Loving Grace.” In it, he predicted that by 2027, AI would be capable of compressing “a hundred years of scientific progress into ten years.” He also spoke, in the same text, of existential risks to humanity.
This is the man entering the White House—a man who believes that his own invention could either save the human race or render it obsolete. And no one knows which side he’s really on.
What's going on behind those closed doors
The upcoming presidential executive order
According to three corroborating sources cited by Bloomberg and Reuters, the Trump administration is preparing an executive order that could transform the industry. On the agenda: priority access to federal land for AI data centers, accelerated environmental deregulation, and above all—this is where it gets complicated—“critical national infrastructure” status for AI companies deemed “aligned with U.S. interests.”
Translation: a handful of companies will receive government privileges. Is Anthropic one of them? The answer may be decided right now, behind the doors Amodei has just walked through.
The Race Against Beijing: The Real Issue
Forget all the talk about “AI ethics.” The real issue is China. DeepSeek, Qwen, Baidu—Chinese models are catching up. Some are already outperforming them on specific benchmarks. Washington is panicking. Silicon Valley is panicking. And Amodei, for his part, has been peddling the same idea for the past eighteen months: the United States must massively subsidize domestic AI or lose the technological war of the century.
He didn’t come to ask for anything. He came to negotiate the terms of an American victory.
And yet, something doesn't feel right
The man who preached caution is now peddling power
Here is the paradox that should haunt every clear-thinking citizen. The man who left OpenAI, accusing the company of neglecting safety, is the same man who is now pushing the U.S. government to lift regulatory restraints on his own industry.
Amodei recently stated that overly strict regulation of AI “would be tantamount to handing the 21st century over to Communist China.” An unassailable statement. An effective statement. A statement that turns any attempt at democratic oversight into an act of treason.
The Shift No One Dares to Name
In three years, Anthropic has gone from an ethics-focused California startup to a major geopolitical player negotiating contracts with the Pentagon, Palantir, and AWS. In November 2024, the company signed a partnership with Palantir—yes, Peter Thiel’s company, the one that supplies surveillance systems to U.S. intelligence agencies.
How does a company founded to “protect humanity from AI” end up providing AI to the agencies that monitor humanity? That’s the question no one is asking Amodei. Because no one is in the room with him today.
The Faceless Power of the New Era
When CEOs Replace Diplomats
Twenty years ago, crucial meetings at the White House were held with heads of state, foreign ministers, and ambassadors. Today, they’re held with Sam Altman, Elon Musk, Jensen Huang, and Mark Zuckerberg. And now, Dario Amodei.
Power has changed hands without anyone voting. The decisions that will shape work, health, education, war, and peace in the 21st century are now made in upholstered lounges, among men in their forties who control models that no one else truly understands.
Opacity as a Strategy
No minutes will be published from this meeting. No public archive will preserve what was said. Tech lobbyists have quietly ensured that their meetings with the president are exempt from the transparency requirements that apply to other industries.
This is no oversight. It is a deliberate design choice by the new administration.
What This Visit Reveals About Our Democracy
The citizen is no longer in the room
When Amodei leaves the White House tonight, he’ll get back into his sedan. He’ll return to his private jet. He’ll head back to San Francisco. And somewhere, in the weeks that follow, an executive order will be signed. Billions of dollars will be allocated. Rules will be written. Lives will be transformed.
No citizen will have been consulted. No parliamentary debate will have taken place. No journalist—I’m not a journalist, I’m a columnist, but even real journalists—will have been able to ask a single question in the room where these decisions are made.
The Manufactured Consent of the AI Era
Meanwhile, on our screens, Anthropic is issuing reassuring press releases. “Constitutional AI.” “Responsible scaling.” “Alignment research.” The vocabulary is that of ethics. The language is that of caution. But the actions themselves are those of a company that is taking over the government while the government believes it is taking over the company.
The Questions No One Asks Amodei
Six simple questions, six deafening silences
Who is speaking in this room? A man who represents his shareholders—Google, Amazon, Spark Capital—and not all of humanity, despite his speech.
Who is absent? The workers whom his models will replace. The artists whose works Claude used without consent. The citizens of the Global South who will suffer the consequences without having a say in the matter.
Who benefits? Anthropic’s shareholders, with the company valued at 180 billion. Elected officials who receive campaign donations. Federal agencies that gain access to superhuman surveillance tools.
What’s been turned on its head? Regulation is portrayed as a danger, when in fact the lack of regulation is the real danger. Speed is presented as a necessity, when it is actually a choice.
Why now? Because the political window is open. Trump wants a technological trophy. Amodei wants a competitive edge over OpenAI and Beijing.
What if it were the other way around? If a Chinese CEO met with Xi Jinping in private to negotiate state privileges regarding AI—how would our media report on it? With the same deference?
For democratic vigilance commensurate with the danger
What We Have Left
We won’t be able to stop Dario Amodei from meeting with Donald Trump. We won’t be able to force the release of the meeting minutes. We won’t be able to reverse the tide of twenty years of regulatory capture by Big Tech.
But there is one thing we can refuse to do. We can refuse the complicit silence surrounding these meetings. We can call out what is happening. We can ask the questions that embedded journalists no longer dare to ask.
The Final Question
Ten years from now, when artificial intelligence has transformed the world beyond recognition, our children will ask us: Where were you on November 18, 2025, the day a man walked alone into the White House to chart the future of our species?
I want to be able to answer: I was watching. I was taking notes. I refused to look away.
An uncompromising verdict
Neither demonization nor naivety
Dario Amodei is no movie villain. He is likely a sincere, anxious, brilliant man who genuinely believes he is acting for the greater good. That is precisely what makes the situation dangerous. Democratic disasters are not engineered by tyrants. They are engineered by men convinced that their superior intelligence justifies their superior power.
What I take away from this day
A black sedan. A gate that opens. A man who disappears into a building. Four seconds of footage. And behind those four seconds, the quiet shift of an entire century toward a model where decisions shaping civilization are made without us, for us, and sometimes against us.
Democracy does not die in a coup d’état. It dies in the boardrooms where cameras are no longer allowed.
Time for Civic Awakening
What Everyone Can Do
Demand transparency regarding meetings between tech executives and government officials. Support the few elected officials who are calling for public hearings with AI CEOs. Read—really read—the terms of service for the tools you use. Talk to your loved ones about these issues with the seriousness they deserve.
And above all: reject the idea that these topics are too complex for you. They aren’t. They’re deliberately complicated, deliberately obscured, deliberately locked away. But they concern you. Directly. Completely. Irrevocably.
The World to Come
Artificial intelligence will be the most powerful technology humanity has ever created. The question is no longer whether it will transform our lives—it already is. The question is who will control this transformation. A handful of men in closed-off offices. Or all of us, together, with our eyes wide open.
Today, Dario Amodei has made his choice. Tomorrow, it will be our turn.
The Last Word Before It's Too Late
Nothing is decided yet
The history of technology always shows the same thing: there is a window of opportunity. A narrow, fragile, temporary window during which citizens can still influence the course of events. That window is open right now for artificial intelligence. It won’t stay open for long.
What I hope for
I hope that one day, in another White House, another president will meet with another AI CEO—with journalists in the room, citizens on the balconies, and Congress serving as a real check on power. I hope this era will be remembered as a tipping point that was narrowly averted by a collective surge of action. I hope we won’t be the generation that let the hurricane pass without even closing the shutters.
Today, a man entered the White House. Tomorrow, he’ll leave with commitments we’ll never know. Between those two moments, there’s us. And our vigilance. And nothing else.
Let’s stay alert.
Signed, Jacques PJ Provost
Transparency Box
What I Know
The facts reported in this article are based on a Reuters dispatch published by The Straits Times on November 18, 2025, confirming Dario Amodei’s arrival at the White House for discussions that have not been publicly specified. The contextual details—the Anthropic-Palantir partnership, Amodei’s public statements, and Anthropic’s valuation—are documented in the company’s official statements and media coverage from recent months.
What I Do Not Know
I do not know the exact content of the discussions between Dario Amodei and the Trump administration. I do not know what agreements, commitments, or contracts were discussed. I do not know who else was present in the room. This information had not been made public at the time of writing.
My role
My role is to interpret these facts, contextualize them within the framework of contemporary geopolitical and technological dynamics, and give them coherent meaning within the broader narrative of the transformations shaping our era. These analyses reflect expertise developed through continuous observation of international affairs and an understanding of the strategic mechanisms that drive global actors.
Any subsequent developments in the situation could, of course, alter the perspectives presented here. This article will be updated if major new official information is released, thereby ensuring the relevance and timeliness of the analysis provided.
Sources
Primary Sources
Anthropic — Official press releases and company news — 2024–2025
Secondary Sources
Reuters — Artificial Intelligence Coverage — 2025
Bloomberg — Special Report on Artificial Intelligence and U.S. Policy — 2025
Machines of Loving Grace — Essay by Dario Amodei — October 2024
This content was created with the help of AI.