OpenAI’s Pentagon Deal Hands Anthropic a Market Opening as Defense Tensions Test AI Strategy
ChatGPT uninstalls surged 295% after Sam Altman's rushed Defense Department agreement, while Claude overtook it in app rankings—a competitive shift with implications for US AI sovereignty and China competition.
OpenAI’s military entanglement triggered the largest user exodus in ChatGPT’s history, sending 1.5 million users toward rival Anthropic and exposing fault lines between commercial AI ambitions and national security imperatives.
ChatGPT app uninstalls jumped 295% on February 28, when OpenAI CEO Sam Altman announced a Pentagon deal hours after President Trump banned federal use of Anthropic’s Claude, according to TechCrunch. Claude surged past ChatGPT to become the most downloaded free app in Apple’s App Store, marking the first time OpenAI has ceded the top position. Since the start of 2026, free active users on Claude have increased by over 60%, and daily sign-ups have quadrupled, according to Fortune.
The controversy stems from Anthropic CEO Dario Amodei refusing the Pentagon’s request for unrestricted access to Claude for mass Surveillance and autonomous weapons, per Euronews. Defense Secretary Pete Hegseth summoned Amodei to the Pentagon, demanding Anthropic drop contractual prohibitions or face designation as a supply-chain risk, according to Understanding AI. When Anthropic held firm, Trump directed federal agencies to immediately cease use of Anthropic’s technology.
The Rushed Deal That Backfired
After negotiations between Anthropic and the Pentagon collapsed Friday, Trump directed federal agencies to stop using Anthropic’s technology, and OpenAI quickly announced it had reached a deal for models to be deployed in classified environments, reports TechCrunch. Altman later admitted he “shouldn’t have rushed” the deal out on Friday, saying “it just looked opportunistic and sloppy”, according to CNBC.
The timing proved disastrous for OpenAI’s public positioning. The deal came right after talks between Anthropic and the Defense Department broke down, though Altman had told employees in a Thursday memo that OpenAI shared the same “red lines” as Anthropic, per CNBC. Many OpenAI employees respect Anthropic for standing up to the Pentagon and felt frustrated when Altman surprised them by saying he agreed with rival CEO Dario Amodei before announcing OpenAI had been negotiating its own deal, reports CNN.
“We were genuinely trying to de-escalate things and avoid a much worse outcome, but I think it just looked opportunistic and sloppy.”
— Sam Altman, OpenAI CEO
Campaign finance reports showed OpenAI President Greg Brockman and his wife each donated $12.5 million to MAGA Inc., making up nearly a quarter of the roughly $102 million it raised in late 2025, according to MIT Technology Review. The $25 million combined donation became a rallying point for the QuitGPT boycott campaign, which claims more than 1.5 million people have taken action, either by cancelling subscriptions, sharing boycott messages, or signing up via quitgpt.org.
Anthropic Capitalizes With Strategic Features
Anthropic moved quickly to capture defecting users. The company launched a memory import tool allowing users to transfer conversations and preferences from ChatGPT and Gemini to Claude, so new users don’t need to start from scratch, reports MacRumors. Anthropic made memory features available on the free plan, removing a previous barrier to switching.
By Monday morning, Claude experienced an outage that the company attributed to “unprecedented demand”, according to Gizmodo, underscoring the scale of user migration.
Defense Secretary Pete Hegseth warned Anthropic CEO Dario Amodei that if the company refused to allow its AI models to be deployed for “all lawful purposes,” including potential applications in surveillance and fully autonomous weapons, the Pentagon would terminate its contract and label Anthropic a national security risk. Amodei rejected the terms outright, according to Fortune.
The stance resonated with consumers prioritizing AI ethics over performance. ChatGPT users protested OpenAI’s Pentagon contract online by spreading “Cancel ChatGPT” messages across Reddit and X, posting guides for deleting accounts and migrating to Claude. Even celebrity support materialized: pop star Katy Perry posted “done” alongside a screenshot of the Claude Pro sign-up page with a heart over it, per Gizmodo.
Defense Contracts and AI Sovereignty Collide
The Pentagon standoff exposes deeper tensions between commercial AI development and national security requirements. OpenAI’s contractual language does not appear to meaningfully restrict the government’s ability to spy on Americans or build fully autonomous weapons, legal experts told Understanding AI. After OpenAI published contract details, Techdirt’s Mike Masnick claimed the deal “absolutely does allow for domestic surveillance,” citing compliance with Executive Order 12333—the legal framework that permits the NSA to capture communications by tapping lines outside the US, even containing info on US persons.
- Anthropic’s Position: Hard contractual bans on mass domestic surveillance and fully autonomous weapons without human oversight
- OpenAI’s Approach: Technical “safety stack” controls plus contract language referencing existing US law and policy
- Pentagon Demand: AI models available for “all lawful purposes” without explicit prohibitions
- Legal Ambiguity: Executive Order 12333 permits intelligence collection on non-US persons that can incidentally capture US citizen data
Former deputy assistant secretary of defense Michael Horowitz said the dispute reflects a breakdown in trust, where Anthropic does not trust that the Pentagon will use their tech responsibly, and the Pentagon doesn’t trust that Anthropic will allow its tech to be used for important national security use cases, according to NBC News.
The broader geopolitical stakes are substantial. Technology is perhaps the most intense realm of competition between the United States and China today, and artificial intelligence is central to that contest. By developing state-of-the-art capabilities in AI, China seeks to achieve a strategic advantage over the US and its allies, according to Brookings Institution. The internal divisions between US AI companies—and between commercial players and defense needs—create vulnerability.
China Competition and Strategic Fragmentation
The Pentagon’s hardline approach stems partly from concerns about Chinese AI advancement. Company leaders acknowledged that governments will spy on adversaries internationally, referencing threat intelligence reports showing China was already using AI models to target dissidents overseas, according to sources at OpenAI’s all-hands meeting reported by Fortune.
| Dimension | US Approach | China Strategy |
|---|---|---|
| Development Model | Private sector innovation, open markets | State-led development, sovereign control |
| Governance Philosophy | Market-driven with allied partnerships | Cyber sovereignty, UN-led mechanisms |
| Data Framework | OECD principles, G7 coordination | Mandatory state access, “secure” flows |
| Export Strategy | Chip restrictions, trusted partner networks | Belt & Road integration, Global South positioning |
| Military Integration | “All lawful purposes” doctrine | Civil-military fusion model |
Middle powers that fail to secure influence over AI development, deployment and governance will likely forfeit control over their economies and political systems. “Sovereign AI” capabilities—preserving autonomy in AI policy choices—are becoming as fundamental to national power as military strength, according to Chatham House.
The OpenAI-Anthropic split undermines US technological cohesion at a critical juncture. Recent moves by Washington and large US tech companies reflect an attempt to compel other countries into a binary choice: will their AI technology be delivered by America or by China? Microsoft President Brad Smith stated the number-one factor defining whether the US or China wins the race is whose technology is most broadly adopted globally, per Chatham House.
The most significant question for OpenAI is why they believe they would be allowed to exercise any objections to the DOD contract after the government attempted to destroy Anthropic for what OpenAI argues were the same or worse contract terms, writes the Center for American Progress. That vulnerability could constrain US AI companies’ ability to refuse future Pentagon demands, potentially driving allies toward Chinese alternatives or sovereign AI programs.
Market Implications and Financial Pressure
The user exodus hits OpenAI during financial strain. OpenAI CEO Sam Altman has put the company $96 billion in debt, according to Black Enterprise. OpenAI closed $110 billion in new funding at an $840 billion post-money valuation on the same day as the Pentagon announcement—the largest private financing in history, reports TheStreet.
Anthropic benefits from positioning as the ethical alternative. Before Super Bowl ads taking a dig at OpenAI’s move to include ads in ChatGPT, Anthropic ranked 42nd on Apple’s App Store. Since then it has remained in the top 10 most downloaded apps, but OpenAI’s Pentagon contract may have helped push it into the top spot, according to Fortune.
Anthropic CEO Dario Amodei has resumed negotiations with the Defense Department to put a military contract back on the table, per the Financial Times, suggesting the standoff may not be permanent. Whether Anthropic can secure acceptable terms while maintaining its ethical positioning—or whether the Pentagon’s “all lawful purposes” doctrine becomes industry standard—will shape AI companies’ leverage going forward.
What to Watch
Congressional oversight looms. Anthropic CEO Dario Amodei should testify before Congress immediately—as should Defense Secretary Hegseth—to help the public understand events between the two parties. OpenAI CEO Sam Altman should testify to explain differences between their DOD contract compared to Anthropic’s demands, argues the Center for American Progress.
Track whether ChatGPT’s user base stabilizes or continues declining—sustained attrition would validate the QuitGPT campaign’s leverage and force OpenAI to renegotiate public positioning. Monitor Claude’s paid conversion rates: free user growth only translates to business advantage if Anthropic can monetize the influx.
Watch allied government responses, particularly in Europe and Five Eyes nations. If key partners adopt restrictive AI procurement policies citing surveillance concerns, it validates Anthropic’s stance and constrains OpenAI’s international expansion. Conversely, if allies follow the Pentagon’s lead, it establishes “all lawful purposes” as the defense industry standard.
The China dimension remains critical. DeepSeek’s R1 model boosted Beijing’s confidence in its ability to compete at the frontier despite US semiconductor restrictions. Developed on Chinese-made AI accelerators and optimized for efficiency, R1 hit performance benchmarks competitive with leading US models while relying on fewer high-end Western chips, according to Geopolitical Monitor. If China continues closing the technical gap while US companies fragment over defense ethics, America’s AI advantage narrows faster than export controls can compensate.
The Pentagon’s willingness to blacklist a major AI lab sets precedent the entire industry must navigate. Whether that pressure forges a unified US approach or drives fragmentation—handing competitive advantage to China or enabling European sovereign AI programs—depends on the next three months of contract negotiations and user retention data.