Pete Hegseth Threatens To Blacklist Anthropic Over National Security Concerns

Good morning!

A little bit of news relating to AI and national security.

Secretary of Defense Pete Hegseth (or as some call it, the "Secretary of War") is close to blacklisting beloved Claude creator Anthropic.

Why?

Anthropic refused to allow Claude to be used for mass surveillance systems in the US and autonomous weapons like drones.

The Pentagon wants "all lawful purposes" access to Claude AI.

No restrictions. No safeguards.

They want to use it for intelligence gathering, surveillance, weapons development, and battlefield operations.

But Anthropic said no to two specific things:

• Mass surveillance of American citizens

• Fully autonomous weapons that operate without human oversight

The Pentagon didn't like that answer.

So now they're threatening to blacklist Anthropic as a "supply chain risk." That's a label usually reserved for foreign adversaries like Chinese tech companies.

This would kill their $200 million contract and force all Department of Defense contractors to stop using Claude entirely.

Why does the Pentagon want unrestricted access to Claude?

According to DoD officials, it's about national security.

Our adversaries (especially China and Russia) are aggressively developing military AI without any ethical constraints.

They're not worried about safeguards. They're racing ahead.

The Pentagon argues that restricting AI capabilities could put American troops at risk and give our enemies an advantage.

Why The Department Of Defense Wants AI:

1. Speed and efficiency

   AI can process intelligence data faster than any human analyst, helping commanders make decisions in real time during combat.

2. Improved accuracy

   AI can analyze patterns and predict threats more reliably, potentially saving lives by identifying dangers before they escalate.

3. 24/7 operations

   Unlike humans, AI never sleeps, never gets tired, and can monitor threats around the clock.

4. Cyber defense

   AI can detect and respond to cyberattacks instantly, protecting critical military systems from foreign hackers.

5. Competitive edge

“If China deploys unrestricted military AI and we don't, we risk falling behind in the global AI arms race” is basically what the Pentagon is saying

The Pentagon believes that ethical red lines could cause AI systems to "refuse" critical tasks mid-operation, potentially compromising missions where every second counts.

In their view, safeguards are a liability when you're trying to win wars.

That said, I understand why Anthropic doesn't want Claude weaponized.

There are serious downsides to using AI the way the Pentagon wants.

1. Risk of accidents

   AI can make mistakes. What happens when an autonomous weapon misidentifies a target and kills innocent civilians? Who's responsible?

2. Escalation of conflicts

   Autonomous weapons could make split-second decisions that escalate situations into full-blown wars before humans can intervene.

3. Erosion of trust

   Mass surveillance of American citizens violates constitutional rights. People lose trust in their government when AI is used to spy on them without warrants.

4. Risk of boycott from everyday users

   If Anthropic agrees to remove these safeguards, everyday users might abandon Claude entirely. Think about it. If people find out Claude is being used for mass surveillance or autonomous weapons, they'll stop trusting the company. They'll switch to other AI tools that haven't compromised their ethics.

5. No accountability

   When AI makes a decision that leads to war crimes or civilian deaths, who do we hold accountable? The programmer? The military commander? The AI itself?

6. Unpredictable behavior

   AI systems can develop emergent behaviors that nobody anticipated. In weapons systems, that's terrifying.

7. Cyberattack vulnerabilities

   Rushing AI into military systems without proper safeguards makes them targets for hacking, data poisoning, or manipulation by foreign actors.

These aren't hypothetical concerns.

We've already seen AI-driven targeting programs increase collateral damage in past conflicts.

The more autonomy we give to AI in warfare, the more we risk unintended consequences that could spiral out of control.

So what happens if Anthropic gets blacklisted?

The Pentagon has backup plans.

OpenAI and xAI are both negotiating similar $200 million contracts with the Department of Defense.

And unlike Anthropic, they're showing "more flexibility" on removing safeguards.

OpenAI has already added ChatGPT to the Pentagon's GenAI.mil platform, giving over 3 million military and civilian users access. They've agreed to lift standard guardrails for Pentagon work and are pushing for classified network expansions.

xAI's Grok was added to military systems in early 2026 and is already integrated for secure information handling. Elon Musk's involvement has accelerated the process, and xAI is even competing in a $100 million drone swarm contest.

Both companies are ready to step in if Anthropic walks away.

So in the end, the DoD will still get what it wants.

There will be a setback (switching AI models takes time and costs money) but it won't stop the Pentagon's push for unrestricted military AI.

Personally, I don't like the idea of AI being weaponized.

We've all seen what happened in The Terminator.

Giving machines the power to make life-or-death decisions without human oversight is just plain dangerous.

But the reality is our adversaries are doing it regardless.

China isn't worrying about ethics. Russia isn't adding safeguards.

I understand why the Department of Defense doesn't want to be the only one playing by the rules while everyone else cheats.

It's a tough call.

So I’m just hoping that whatever they do with AI, it’s done with the people’s best interest in mind.

If we're going to use AI in military applications, we need strong oversight, accountability, and transparency.

Not just a free-for-all where AI does whatever it wants.

What do you think?

Should the Pentagon have unrestricted access to AI for national security?

Or should companies like Anthropic hold firm on ethical safeguards, even if it means losing government contracts?

A quick sidenote

If the government is using AI for their day-to-day operations, there's no reason for you not to be using it in your business.

Whether it's automating emails, scheduling appointments, or analyzing data, AI can save you time and money right now.

Don't wait for permission. Start using it today.

Join us in the next AI For Business Summit

Talk soon,

Brian