The number of uninstalls of the chatbot app surged by 295%.

OpenAI will amend its agreement with the Pentagon following a wave of criticism and a mass exodus of users to Anthropic. This was announced by the CEO of the AI startup, Sam Altman.

Here is a re-post of an internal post:

We have been working with the DoW to make some additions in our agreement to make our principles very clear.

1. We are going to amend our deal to add this language, in addition to everything else:

"• Consistent with applicable laws,…

— Sam Altman (@sama) March 3, 2026

"One thing I did wrong: we shouldn't have rushed into the deal. The issues are very complex and require clear communication. We genuinely tried to reduce tension and avoid a much worse outcome, but I think it came off as opportunistic and sloppy," he wrote.

OpenAI signed a contract with the U.S. Department of Defense on February 28. Just hours before, President Donald Trump ordered federal agencies to stop using Anthropic's tools.

This decision followed a statement from the project's CEO, Dario Amodei, who publicly prohibited the use of the developer's neural networks for mass surveillance and the creation of autonomous weapons.

New Terms

In the revised contract, OpenAI added a clause prohibiting the "intentional use of artificial intelligence" for spying on U.S. citizens. It also specifies that restrictions apply to the purchase or use of commercial personal data databases.

The U.S. Department of Defense also confirmed that the company's tools will not be used by intelligence agencies, including the NSA. Any supplies to such agencies will require additional agreements.

"It is extremely important to protect the civil liberties of Americans. This has received so much attention that we wanted to clarify this point particularly clearly," explained OpenAI's CEO.

He emphasized that democratic processes must be followed in controlling the development of AI. According to Altman, no private company should determine the fate of the world.

The CEO of ChatGPT urged the Pentagon not to list Anthropic as a threat to supply chains and to offer competitors the same terms as OpenAI.

"I think the real dangers to the world are still ahead. I tried to put myself in the shoes of someone who wakes up the day after an attack on the U.S. or the emergence of new biological weapons that we could help prevent," Altman concluded.

ChatGPT Boycott

Following the news of the Pentagon deal, American users began to uninstall ChatGPT en masse, according to TechCrunch. On March 1, the number of app uninstalls increased by 295%, as reported by Sensor Tower.

The average daily uninstall rate over the previous 30 days was 9%, analysts noted.

The trend in new downloads shifted to a decline: immediately after the agreement, ChatGPT downloads plummeted by 13%, and the following day by another 5%. The app's rating also dropped significantly: the number of five-star reviews halved, while negative reviews surged 8.5 times on the first night and doubled on March 2.

ChatGPT is getting COOKED on the iOS App Store reviews and I think it’s deserved. pic.twitter.com/QaCZx7iHrC

— Not Greg (@dogecoinmillion) March 3, 2026

Downloads of Claude from Anthropic soared by 37% on February 28 and another 51% on March 1. The chatbot rose to the top of the U.S. App Store and maintained its lead as of March 3.

Other analytical platforms confirmed this trend. Appfigures reported that daily installations of Claude in the U.S. surpassed ChatGPT for the first time: on March 1, the figure increased by 88%. The app also topped the free iPhone app rankings in Belgium, Canada, Germany, Luxembourg, Norway, and Switzerland.

Similarweb calculated that downloads of Claude in the United States over the past week exceeded January's figures by about 20 times. However, experts cautioned against jumping to conclusions, suggesting that other factors may have influenced the statistics.

In addition to abandoning ChatGPT, over 700 employees from Google and OpenAI also signed an open letter. In the document, they called for a ban on the use of AI in surveillance systems or for combat tasks without direct human oversight.

Recall that in mid-February, Bloomberg journalists reported on SpaceX's plans to create software for the Pentagon's autonomous weapons.