OpenAI Employees Resign To Protect You From AI Surveillance

Good morning!

Some hot news right now about the future of AI and National Security.

More of OpenAI's senior employees just resigned.

This came after the company signed a deal with the Pentagon that lets the government use ChatGPT and other OpenAI models for autonomous weapons and mass surveillance of Americans.

Anthropic Got Blacklisted For Refusing The Same Thing

You may remember that I wrote a newsletter a few weeks ago about how the Pentagon wanted to use Claude AI for military purposes.

Anthropic said no.

They refused to let Claude be used for mass surveillance of American citizens and Autonomous weapons that operate without human oversight

The Pentagon didn't like that answer.

So they threatened to blacklist Anthropic as a "supply chain risk". It’s a label usually saved for foreign enemies like Chinese tech companies.

Anthropic stood firm. They got blacklisted.

And OpenAI swooped in and took the $200 million+ deal instead.

Quick Recap: Why Does The Pentagon Want AI - And What Does This Actually Look Like?

The ‘Department of War’ has some pretty clear reasons for wanting unrestricted access to AI to have a military edge against Russia and China who are developing AI without guardrails.

The Pentagon's basic argument is this

If we tie our hands with ethics while our enemies don't, we risk putting American troops in danger and losing the global AI arms race.

In their view, safeguards could cause AI to "refuse" critical tasks during operations when every second counts.

But Here's What The Government Having “Unrestricted Access“ To AI Means For Regular Americans

1. Mass Surveillance Without Warrants

   The Pentagon can legally skip getting a warrant by simply buying your personal data from commercial data brokers.Think about all the apps on your phone - fitness trackers, shopping apps, social media, location services. Many of them sell your data to third parties. The government calls this "publicly available information,"

   Which means they can buy it and use AI to analyze:

• Where you go every day

• Who you talk to and when

• What you buy and search for online

• Your political views and personal beliefs

• Your health conditions and habits

All without a judge ever signing off on it.

AI can then connect all these dots to build a detailed profile of you - and millions of other Americans - to predict behaviors and identify "threats."

No warrant needed. No human reviewing each case. Just AI making decisions about who might be dangerous.

2. Autonomous Weapons In Action

   While the Pentagon says it has "no intention" of deploying fully autonomous weapons, they've refused to sign any contractual ban on them.

   And they're already using AI for real combat decisions.

   In January 2026, AI was reportedly used for "real-time analysis" during the operation that captured Venezuelan leader Nicolás Maduro.

The Backlash? OpenAI Is Facing A Mass Exodus Of Senior Employees And Millions Of Users

Last time, I warned about the backlash and loss of trust if Anthropic agrees to the deal. They dodged a bullet, and instead, OpenAI was the one that got hit.

The backlash has been massive.

1. Over 100 current OpenAI employees signed a public letter protesting the Pentagon deal. They're supporting Anthropic's ethical stance and telling their own bosses to add the same safeguards.

2. Caitlin Kalinowski, a senior leader in OpenAI's robotics division, resigned on March 7, 2026. She said the Pentagon deal was "rushed without the guardrails defined." Her resignation post got over 7 million views and 58,000 likes.She's not the only one. Other top researchers have quietly left too.

3. Over 1.5 million users pledged to cancel ChatGPT. App uninstalls spiked. Claude jumped to #1 in the App Store as people switched away from OpenAI.

Even CEO Sam Altman admitted the deal was "opportunistic and sloppy." He's now scrambling to add clearer bans on domestic spying and autonomous weapons.

It's Easy To See Why People Are Upset

We're talking about autonomous AI weapons and surveillance systems with little to no human oversight.

This isn't science fiction anymore. This is real technology being deployed right now with real consequences.

When AI makes a mistake in combat, innocent people pay the price. And when things go wrong, there's no clear person to hold accountable.

How These Events Play Out Will Dictate The Future Of Our Country And Pretty Much The World

If this deal goes through without strong safeguards, it sends a clear signal to the entire world:

“The leading free and democratic country on earth has allowed the development of AI mass surveillance on its own citizens.”

Just like how Russia does it. Just like how China does it.

We'd be telling every other country that it's okay to build AI weapons that operate without human authorization and spy on people without judicial oversight.

This is a turning point. The decisions being made right now

About warrants, human oversight, and autonomous weapons

It will shape AI development for decades.

Let Me Know What You Think

I want to hear from you on this one.

Do you think the Pentagon should have unrestricted access to AI for national security purposes?

Or should all AI companies follow Anthropic’s example and hold firm on ethical safeguards, even if it means losing government contracts?

Talk soon,

Brian