The US Department of Defense has awarded major contracts, each capped at $200 million, to OpenAI, Google, Anthropic, and Elon Musk’s xAI as part of a vast effort to integrate advanced AI tools into military systems. These deals focus on developing so-called “agentic AI workflows,” which are designed not only to analyze information but also to execute multi-step tasks autonomously.

For OpenAI, this follows a previous collaboration with defense contractor Anduril and signals a full embrace of AI in national security mission areas. The company remains committed to its core principle that its technology won’t be used for offensive weapons. Google, Anthropic, and xAI are similarly working on AI systems to provide capabilities in intelligence, operations and enterprise logistics, with xAI launching its Grok for Government suite to serve across federal agencies.

These contracts reflect a broader shift in the Pentagon’s strategy: relying on cutting-edge commercial AI talent and tools rather than custom-built defense systems. The Department’s goal is to maintain strategic advantage by embedding intelligence into everything from battlefield decision-making to back-office systems.

But these moves haven’t come without criticism. After Grok produced antisemitic output, questions were raised about the readiness and safety of deploying such systems in mission-critical settings. Critics are urging stricter guardrails and oversight, warning that trust in AI systems must be earned, especially when lives are at stake.

What this means going forward:

• Governments see frontier AI not as a luxury, but as essential to future defense strategies.

• Commercial AI companies are rapidly becoming key national security partners.

• Ethical and safety frameworks will need to evolve quickly to keep pace with deployment.

• The success of these programs hinges on combining powerful AI with accountability, human oversight, and rigorous testing.