OpenAI goes on defense as Anthropic surges after Pentagon fallout 

OpenAI is in the hot seat this week over the artificial intelligence company’s new deal with the Pentagon, struck just hours after the agency’s negotiations with competitor Anthropic over safety guardrails fell apart.

The response was nearly immediate, with uninstalls of its flagship ChatGPT app rising 295 percent day-over-day last Saturday, according to reports citing market intelligence provider Sensor Tower. Meanwhile, Anthropic’s Claude app hit No. 1 in the App Store as users flocked to the app in a possible sign of support.

By Monday evening, OpenAI CEO Sam Altman shared an internal post to X detailing new additions to the Pentagon agreements to make the company’s “principles very clear.” In doing so, the company’s co-founder acknowledged the company “shouldn’t have rushed to get this out” last Friday.  

“The issues are super complex and demand clear communication,” Altman wrote. “We were genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy.” 

While Altman called it a “good learning experience,” it is unclear whether the admission will placate critics or rebuild trust following the backlash. Multiple users on X responded to Altman’s post by requesting the contract itself to be released in a show of transparency.  

In one response to Altman’s post with 13,000 views, one user said the “only way” to “regain any trust” is to release the contract document itself, writing, “You guys completely torched your brand and integrity on this.”  

Former OpenAI safety researcher Steven Adley wrote on X that the company “wants you to just trust them that the NSA is excluded from their contract,” stating he hopes it is “clear why, without strong evidence to the contrary, people are mistrusting OpenAI ousting OpenAI on this.” 

The spat has gotten the attention of lawmakers on Capitol Hill, with Democratic Sen. Brian Schatz (Hawaii) posting to X Tuesday that he “just downloaded Claude.” The issue could also be the beginning of a legislative debate.  

Silicon Valley Rep. Sam Liccardo (D-Calif.) said on Monday he will introduce an amendment to the Defense Production Act this week to prohibit federal agencies from “retaliating” against high-risk technology vendors and developers that try to limit the deployment of their technology in “ways to mitigate the risk to United States citizens.” And Sen. Ron Wyden (D-Ore.) pledged to fight the actions against Anthropic, Bloomberg reported.  

OpenAI’s amended agreement now includes language that is “consistent with applicable laws,” Altman said, including the declaration that the “AI system shall not be intentionally used for domestic surveillance of U.S. persons and nationals.” 

“For the avoidance of doubt, the Department understands this limitation to prohibit deliberate tracking, surveillance, or monitoring of U.S. persons or nationals, including through the procurement or use of commercially acquired personal or identifiable information,” Altman said, adding it is “critical to protect the civil liberties of Americans.”  

The Pentagon, according to Altman, also affirmed to OpenAI that its services will not be used by the department’s intelligence agencies, including the National Security Agency.  

Altman further defended the deal in an all-hands meeting Tuesday, the Wall Street Journal reported, telling employees he “feel[s] terrible for subjecting” them to the backlash. WSJ reported that Altman described the situation as “really painful,” while stating it was a “complex but the right decision with extremely difficult brand consequences and very negative PR for us in the short term.”  

Katrina Mulligan, OpenAI’s head of national security partnerships, further defended the changes in a lengthy back-and-forth on X. Mulligan said the new agreement gives other AI labs “a better starting place on the issues. When asked for contract language, Mulligan said she does not agree she is obligated to share it, reiterating the intelligence agencies are not a part of the deal.  

Surveillance and tracking were notably among the concerns at Anthropic, which pressed for specific restrictions on mass domestic surveillance and fully autonomous lethal weapons. The Department of Defense wanted language to permit the use of Anthropic’s technology for “all lawful purposes.”  

Following the expiration of the deadline set forth by the Pentagon, Defense Secretary Pete Hegseth announced the Pentagon would label Anthropic as a supply chain risk and President Trump also ordered federal agencies to stop using Anthropic’s technology.  

Anthropic has provided its AI models to U.S. defense and civilian agencies since late 2024 through a partnership with longtime government contractor Palantir, which has faced its own backlash for its work on immigration enforcement. Anthropic said it plans to challenge the supply risk designation in court.  

Copyright 2026 Nexstar Media Inc. All rights reserved. This material may not be published, broadcast, rewritten, or redistributed.

Noem fends off attacks from left and right in heated hearing: 5 takeaways

GOP Sen. Thom Tillis unleashes on Noem: ‘Time after time, I’ve been ...

‘Breathtaking’: Independent senator deplores Rubio’s explanation for Iran ...

Live updates: Trump says US will escort tankers through Strait of ...

Noem faces GOP heat over $220M ad that boosted ‘your name recognition’

Trump knocks Tucker Carlson, Megyn Kelly over Iran criticism: ‘MAGA is ...

Speaker Johnson pushes against war powers resolution: ‘Frightening’

John Bolton says Hegseth needs ‘attitude adjustment’ after Iran briefing

Texas voters set to deliver verdict in competitive Senate primaries 

TSA moves to center of shutdown drama as jittery lawmakers offer warnings for ...

Zinke announces he won’t seek reelection

OpenAI goes on defense as Anthropic surges after Pentagon fallout 

Senate Republicans warn Trump about expanding Iran mission as death toll rises

Minnesota launches investigation that could lead to charges against immigration ...

Pentagon stuns Silicon Valley with Anthropic ban

Early turnout in Texas Democratic primary outpaces GOP as Election Day kicks off

Nancy Mace under investigation by House Ethics Committee

Key GOP senator criticizes Pentagon strategy in Colby hearing 

© The Hill