If Both Companies Drew the Same Lines, Why Did Only One Get Punished?
I use Claude every single day.
It's made by Anthropic — the AI company that, as of last Friday, has been banned from working with the United States government, designated a "supply chain risk to national security," and blacklisted from every defense contractor in the country.
I use it to write strategy documents for my clients. I use it to build consulting frameworks. I use it to research, draft, analyze, and think. It is, without exaggeration, the most important tool in my business.
And last week, the company that built it was treated like a foreign adversary — not because it failed the military, but because it refused to let its technology be used for mass surveillance of American citizens and fully autonomous weapons.
I need you to understand what actually happened. Not the headline version. Not the political spin. What actually happened, in sequence, with receipts.
Because if you run a business, lead an organization, or make decisions about AI — this story is about you, whether you realize it yet or not.
Here's What Happened. In Order.
In June 2024, Anthropic published its usage policy. Two restrictions were explicit: no mass domestic surveillance, no autonomous weapons designed to cause harm. These weren't hidden in fine print. They were the terms of doing business.
In July 2025, the Pentagon signed a $200 million contract with Anthropic anyway — with those restrictions in place. The Department of Defense agreed to them. Claude became the first commercial AI model deployed on the military's classified networks, integrated through a partnership with Palantir. It was used at national nuclear laboratories. It did intelligence analysis directly for the Department of Defense.
Then, in early 2026, the Pentagon told Anthropic to remove those restrictions. Not modify them. Remove them. Defense Secretary Pete Hegseth demanded an AI model "free from usage policy constraints that may limit lawful military applications." When Anthropic's CEO Dario Amodei met with Hegseth, the message was clear: comply, or be designated a supply chain risk and potentially nationalized under the Defense Production Act — a Korean War-era law designed for steel mills and tank factories.
Anthropic said no.
On Friday, February 27th, President Trump posted on Truth Social directing every federal agency to "immediately cease" all use of Anthropic's technology. He called the company's leadership "leftwing nut jobs" and threatened "major civil and criminal consequences."
Hours later, Hegseth designated Anthropic a supply chain risk to national security. This is the first time that designation has ever been used against an American company. It is typically reserved for foreign adversaries. It means every contractor, supplier, and partner working with the U.S. military is now prohibited from doing business with Anthropic.
That same Friday evening — within hours — OpenAI CEO Sam Altman announced that his company had reached a deal with the Pentagon to deploy its models in the same classified networks.
Same Lines. Different Outcome.
Here's where this stops being a political story and becomes something every business leader needs to pay attention to.
OpenAI's deal with the Pentagon includes prohibitions on mass domestic surveillance and autonomous weapons — the exact same red lines that Anthropic was punished for insisting on. OpenAI actually added a third restriction: no "high-stakes automated decisions" like social credit systems.
Read that again. The government designated Anthropic a national security threat for holding positions that OpenAI was rewarded for adopting.
Altman himself told CNBC that his company shares the same "red lines" as Anthropic. In an internal memo to employees, he said OpenAI was seeking to negotiate exclusions preventing use for surveillance in the U.S. — the very thing Anthropic had been demanding.
The difference wasn't the policy. It was the politics.
Follow the Money
Dario Amodei endorsed Kamala Harris before the 2024 election. He donated over $214,000 to Democratic candidates and committees. He once called Trump "a serious and legitimate threat to the rule of law" in a now-deleted Facebook post. Anthropic's backers and board members have collectively given more than $174 million to Democrats.
Sam Altman donated $1 million personally to Trump's inaugural fund. This was after spending years donating to Democrats — including $200,000 to Biden's 2024 reelection campaign. He once compared Trump to Hitler during the 2016 campaign. But by January 2025, he was standing behind the presidential seal at the White House, praising Trump for the $500 billion Stargate AI infrastructure initiative.
OpenAI's cofounder Greg Brockman and his wife committed $50 million to a new AI political venture.
I am not telling you who to vote for. I am telling you to look at the sequence. An AI company that donated to Democrats held a principled position and got blacklisted. An AI company whose CEO donated $1 million to the incoming president adopted the same position and got the contract. On the same day.
If the red lines were the problem, both companies would have been rejected. They weren't. So the red lines weren't the problem.
The Part Nobody Is Talking About
Here's what keeps me up at night about this, and it has nothing to do with which party is in power.
We are watching, in real time, the establishment of a precedent: the U.S. government will punish private companies for maintaining ethical boundaries on their own technology.
This is not about Anthropic vs. OpenAI. This is not about Democrats vs. Republicans. This is about what happens when a government decides that a company's ethics policy is a national security threat.
The Defense Production Act — the tool the Pentagon threatened to use — was designed to compel factories to produce steel and ammunition during the Korean War. It is now being discussed as a mechanism to force an AI company to remove safety guardrails from technology that can surveil millions of citizens simultaneously.
That should concern you regardless of who you voted for.
Because the question isn't whether you trust this administration with that power. The question is whether you trust every future administration with the precedent that this one is setting.
Mass domestic surveillance wasn't illegal before AI. It was just impractical. As Amodei pointed out in a CBS interview, AI makes it possible at a scale that existing law never anticipated. "Domestic mass surveillance is getting ahead of the law," he said. "The technology's advancing so fast that it's out of step with the law."
And instead of asking Congress to catch up, the government punished the company that pointed that out.
What This Means If You Run a Business
I consult with business leaders and nonprofit executives on AI adoption every week. And I know what some of you are thinking: this is a defense contract dispute. It doesn't affect me.
It does. Here's how.
If you use AI tools, you are affected. Every AI company just watched Anthropic get blacklisted for maintaining an ethics policy. The message to every other AI company is clear: don't draw lines the government might not like, or your business could be destroyed overnight. That pressure flows downhill to the tools you use, the data protections you rely on, and the policies that protect your customers.
If you're a government contractor — or ever want to be one — you are directly affected. The supply chain risk designation means that any company doing business with the Pentagon must now certify they don't use Anthropic's models. Fortune 500 general counsels are already asking whether using Claude is worth the risk.
If you care about how AI handles your data, you are affected. The company that drew the hardest line on protecting Americans from surveillance was the one that got punished. What does that tell the next company considering where to draw their line?
My Position
I use Claude. I will continue to use Claude. Not because I'm loyal to a brand, but because Anthropic's position on this is the right one, and I will not be intimidated out of that belief by a political dispute that has nothing to do with national security and everything to do with power.
I believe AI should serve humans. I believe mass surveillance of American citizens is wrong regardless of which party controls the White House. I believe fully autonomous weapons without human oversight are a line we are not ready to cross — not because of ideology, but because the technology is not reliable enough to make life-and-death decisions without a human in the loop. Amodei himself said Anthropic is "not categorically against fully autonomous weapons" but that today's models aren't ready. That's not ideology. That's engineering.
And I believe that when a government punishes a company for holding those positions — while rewarding a competitor for holding the same positions with better political connections — we are no longer talking about national security. We are talking about compliance as a loyalty test.
That's not how democracies work. That's not how free markets work. And that's not how we build technology that serves humanity instead of controlling it.
What Happens Next
Anthropic will challenge the supply chain risk designation in court. Legal experts say the case raises profound constitutional questions about the government's relationship with private enterprise. The statute requires that the Pentagon exhaust less intrusive alternatives before making such a designation — and labeling an American company a national security threat within days of a contract negotiation doesn't look like exhaustion. It looks like retaliation.
In the meantime, Anthropic's seven cofounders have pledged to donate 80% of their wealth. CEO Amodei published a 20,000-word essay warning about the dangers of wealth concentration in AI. The company has been leaning into its safety commitments at the exact moment when backing down would be the easier business decision.
That tells me something about their character. And it tells me something about the character of the moment we're in.
The Question That Matters
This is not a technology story. This is a story about what happens when power meets principle.
Every leader reading this will face a version of this choice — not at the Pentagon's scale, but in your own business. A client will ask you to do something that crosses a line. A vendor will pressure you to compromise a standard. A market will reward the company that cuts corners over the one that holds the line.
When that happens, you get to decide what kind of leader you are. And you get to decide what kind of industry we build.
Anthropic said no. They may pay an enormous price for it. But they will not be the ones who have to explain, years from now, what they were willing to look the other way on.
The companies that say no are not the threat. The moment we start treating them like one — that's the threat.
