OpenAI revises Pentagon deal after backlash as ChatGPT uninstall surge hits the company

OpenAI has amended its recently signed agreement with the US Department of Defense, CEO Sam Altman said, following widespread criticism that the original contract appeared hasty and unclear on key ethical safeguards.

The revision comes amid strong reactions from users, including a significant surge in ChatGPT uninstalls, after the company’s deal with the Pentagon drew backlash across the tech community and among its user base.

Altman acknowledged on social media platform X that OpenAI’s decision to quickly finalise the military contract after rival AI company Anthropic ceased negotiations may have seemed “opportunistic and sloppy.”

He said the company should have taken more time before sealing the pact with the Pentagon, which followed executive action that halted Anthropic’s federal contracts.

As part of the contract revision, OpenAI has clarified the company’s core principles to ensure its AI will not be used by certain intelligence agencies without additional agreement and to reinforce ethical boundaries around the technology’s use.

The original Pentagon deal, announced days earlier, was intended to allow OpenAI’s advanced models to be deployed on classified defence networks, with Altman stating that the company would work with US authorities to embed safeguards such as prohibitions on domestic mass surveillance and ensuring human oversight in any use of force.

However, the announcement triggered widespread criticism both externally and within OpenAI. A research scientist at the firm publicly described the deal as unjustified, and a number of users began switching away from ChatGPT to alternative AI platforms in protest.

Data from analytics firms showed that the ChatGPT app experienced a dramatic spike in uninstalls in the United States, with uninstall rates increasing by nearly 300 percent after news of the Pentagon agreement emerged.

The trend reflected growing discomfort among segments of its user base over the firm’s deeper involvement in military applications.

The backlash has also boosted interest in competing AI services, with rival platforms such as Anthropic’s Claude climbing in popularity and overtaking ChatGPT in app store rankings as some users seek alternatives with stronger safety and ethical guardrails.

Altman responded to the criticism by saying OpenAI has updated its contract principles and will continue to engage with stakeholders on ethical AI deployment, describing the experience as “a good learning moment” for the company as AI technology plays an increasingly central role in both civilian and defence sectors.