Story perspectives

DoD Signs AI Contracts with Tech Giants, Demands Oversight

5/3/2026

Summary

Pentagon Signs AI

May 1, 2026, the U.S. Department of Defense signed contracts with Google, Microsoft, Amazon Web Services, Nvidia, OpenAI, Reflection and SpaceX.
Defense Secretary Pete Hegseth said the deals require lawful Pentagon use and human oversight of autonomous missions.
One contract mandates AI respect constitutional rights and civil liberties, a condition Anthropic previously demanded.
Georgetown University interim executive director Helen Toner warned that AI over-reliance could create automation bias and urged trained operators.

Summary

**Tech Giants Compete**

OpenAI sealed a Pentagon AI deal at end of February, formalized in May.
Hundreds of Google employees, including DeepMind staff, urged CEO Sundar Pichai to stop government AI work.
SpaceX, parent of Elon Musk’s xAI, will supply Grok chatbot for military use.
Nvidia will provide its open-source Nemotron model, with no hardware contract.
Oracle says its cloud lets the “Department of War” build any AI model without vendor lock-in.

Summary

Trump Engages Pentagon

President Donald Trump said his administration had good talks with Anthropic and a Pentagon deal is possible.
Anthropic filed lawsuits against the Defense Department after being labeled a supply-chain risk.
The National Security Agency is testing Anthropic’s Mythos model for software flaws.
600 Google employees signed urging CEO Sundar Pichai to block Pentagon AI work.
Nvidia CEO Jensen Huang said open-source AI models boost safety because transparency aids customization.

Summary

Pentagon Funds AI

The One Big Beautiful Bill Act gave the Pentagon a large AI and offensive cyber budget, prompting tech-firm competition.
The White House resumed talks with Anthropic after its Mythos tool breakthrough, despite the company’s earlier “supply chain risk” label.
A California federal judge blocked the blacklist attempt on Anthropic, while the Pentagon reported that 1.3 million DoD personnel have used GenAI.mil.

History

Historical Context

In 1993, the U.S. Department of Defense launched the Strategic Computing Initiative, a $1.2 billion program to embed artificial intelligence in weapons. IBM contributed about 30 percent of the $1.2 billion budget, and Lockheed Martin produced twelve prototypes that never entered service for the program. The May 2026 Pentagon contracts with Google, Microsoft, Amazon Web Services, Nvidia, and OpenAI echo the SCI’s ambition to embed AI, prompting protests and Anthropic lawsuits. The $10 billion AI procurement budget warns that without stringent oversight, the program may repeat past overruns, shaping future defense policy and civilian-military AI governance.

1 of 5

Pentagon Signs AI

May 1, 2026, the U.S. Department of Defense signed contracts with Google, Microsoft, Amazon Web Services, Nvidia, OpenAI, Reflection and SpaceX.
Defense Secretary Pete Hegseth said the deals require lawful Pentagon use and human oversight of autonomous missions.
One contract mandates AI respect constitutional rights and civil liberties, a condition Anthropic previously demanded.
Georgetown University interim executive director Helen Toner warned that AI over-reliance could create automation bias and urged trained operators.

1 / 5