A dispute between leading artificial intelligence companies and the U.S. government has intensified debate over the role of AI in military operations, after contract negotiations between the United States Department of Defense and AI firms took sharply different paths in early 2026.

The controversy centers on competing deals involving Anthropic and OpenAI, both of which were negotiating contracts worth roughly $200 million to supply AI systems for classified government use.

Talks between the Pentagon and Anthropic collapsed after disagreements over safeguards governing how AI could be used in military and intelligence contexts. Hours later, OpenAI announced a separate agreement with the Defense Department, triggering criticism from lawmakers, AI researchers, and employees within the technology sector.

Negotiations Break Down

Negotiations began in mid-February 2026 as the Pentagon sought to integrate advanced AI tools into classified systems.

According to reporting and company statements, Anthropic insisted on strict prohibitions on two uses of its models:

• Mass domestic surveillance of U.S. citizens
• Fully autonomous lethal weapons without human oversight

Anthropic argued that these restrictions were necessary to prevent misuse of powerful AI systems.

However, U.S. officials reportedly demanded broader authority to use AI tools for “all lawful purposes” related to national security.

The negotiations collapsed on February 27.

Shortly afterward, President Donald Trump ordered federal agencies to cease using Anthropic technology, citing national security concerns. Defense Secretary Pete Hegseth described the company as a “supply chain risk,” an unusually severe designation for a domestic technology firm.

OpenAI Secures Pentagon Deal

Within hours of Anthropic’s blacklist designation, OpenAI announced its own agreement with the Pentagon.

The deal allows deployment of OpenAI models within classified systems under certain safeguards.

OpenAI CEO Sam Altman said the arrangement prohibits the use of AI for mass domestic surveillance and requires human oversight in decisions involving the use of force.

However, critics initially argued that the safeguards relied primarily on existing U.S. law rather than explicit contractual bans.

Following public scrutiny, the agreement was amended in early March to add more explicit restrictions on intentional domestic surveillance and intelligence use without additional approvals.

Industry Divide Over AI Ethics

The dispute highlights growing tensions within the AI industry over how companies should engage with military institutions.

Anthropic has positioned itself as a leader in AI safety, promoting what it calls “constitutional AI” — a framework designed to embed ethical rules directly into AI systems.

The company was founded in 2021 by former OpenAI researchers who left over disagreements about safety governance and commercialization strategies.

OpenAI, meanwhile, has taken a more pragmatic approach, arguing that responsible collaboration with governments is necessary to ensure AI systems are deployed safely.

Altman has warned that AI technologies could be misused in areas such as cyberattacks or biological threats if governments lack access to advanced systems.

Political and Regulatory Fallout

The episode has drawn attention on Capitol Hill.

Senator Ron Wyden criticized the federal blacklist designation against Anthropic and called for closer scrutiny of government influence over AI firms.

Some policymakers and experts have also renewed calls for international agreements governing military uses of AI, including proposals for United Nations discussions on autonomous weapons and AI-enabled warfare.

A New Phase in AI Militarization

The dispute comes amid intensifying global competition in artificial intelligence, particularly between the United States and China.

Governments increasingly view AI as a strategic technology with implications for intelligence gathering, cybersecurity, logistics, and battlefield decision-making.

At the same time, critics warn that the rapid militarization of AI systems raises profound ethical questions about surveillance, accountability, and the future of warfare.

The Anthropic–OpenAI dispute illustrates the difficult balance facing technology companies: whether to impose strict ethical limits on how their systems are used, or collaborate more flexibly with governments pursuing national security objectives.

As AI capabilities continue to advance, similar conflicts between corporate ethics and geopolitical realities are likely to become more frequent.

‍