OpenAI Took a Pentagon Deal, OpenAI signed a deal with the US Pentagon — and the internet revolted. Over 2.5 million users pledged to quit ChatGPT, Claude hit #1 on the weApp Store, and even OpenAI’s own employees pushed back hard.
OpenAI made a deal with the Pentagon on a Friday afternoon. By Monday morning, the internet was on fire.
On February 28, OpenAI announced it had struck an agreement with the US Department of Defense — letting the military deploy its AI models inside classified networks for, in their words, “any lawful purpose.” That phrase did it. Two words. And everything unraveled.
ChatGPT uninstalls in the US surged 295% day-over-day. One-star reviews climbed 775% above baseline levels on both the App Store and Google Play. KQ2 People weren’t just annoyed — they were documenting their exits publicly and handing others a roadmap to leave.
What actually happened with the deal
The drama started weeks earlier. Anthropic — OpenAI’s biggest rival — had been in talks with the Pentagon and walked away. Anthropic CEO Dario Amodei said he “cannot in good conscience” agree to the Pentagon’s request for unrestricted access, citing concerns about autonomous weapons and democratic values. KRDO The Pentagon reportedly blacklisted Anthropic as a supply chain risk over that refusal.
Then OpenAI stepped in and took the contract. Hours after Anthropic got blacklisted.
Sam Altman later admitted the timing was a mistake — saying it “looked opportunistic and sloppy” and that he “shouldn’t have rushed” to get the deal out. Yanko Design He amended the agreement days later to explicitly bar domestic surveillance of US citizens. Critics said the revision didn’t go far enough, and questioned whether the added language was even enforceable.
The fallout got messy
OpenAI’s robotics chief Caitlin Kalinowski resigned over the deal. Protesters gathered outside OpenAI’s San Francisco HQ. Claude shot to the number one spot on the US Apple App Store — and stayed there. GSMArena
On Reddit, detailed migration guides went viral — showing users how to export their ChatGPT history and switch platforms. Posts accumulated tens of thousands of upvotes within 24 hours. KQ2 The boycott had become a movement.
The QuitGPT campaign pointed users toward Claude, Gemini, and open-source alternatives — and strongly advised against switching to Grok. KRDO
Here’s the honest read
2.5 million pledges sounds massive. But ChatGPT has around 900 million weekly users. A 0.3% bleed — even a noisy one — won’t tank OpenAI’s business. What it does signal is something more interesting: people are paying real attention to what AI companies do with their technology. That wasn’t true two years ago.
The QuitGPT website also listed other reasons for distrust — including political donations by OpenAI executives and allegations that a resume-screening tool used by ICE runs on GPT-4. Gizmochina The Pentagon deal was the trigger, not the whole story.
Anthropic didn’t win this by being perfect. It won by refusing one contract at exactly the right moment. That’s a different thing. And it’s worth keeping in mind as Claude becomes the “ethical” AI by default — that reputation now carries enormous pressure to stay that way.
SOURCE: Euronews | NBC News | CNBC
Visit More pages of our site : Write For Us Disclaimer