Story perspectives

Oracle Becomes 8th AI Firm in Pentagon Classified Network.

5/2/2026

Center

Oracle Joins Pentagon

Oracle became the eighth AI company to join the Pentagon’s classified-network agreements, expanding the original list of seven firms.
May 1 2026, Department of Defense signed agreements with AI firms to deploy models on Impact Level 6 and 7 networks for lawful use.
Anthropic was excluded after the Pentagon labeled it a supply-chain risk following a safety-guardrails dispute.
Officials said tools will streamline data synthesis, augment decisions, using GenAI.mil accessed by 1.3 million.

Left

Pentagon AI Contracts

Rights groups claim Pentagon AI contracts could facilitate immigration-raid surveillance and increase civilian casualties in the U.S.–Israel conflict in Iran.
DoD signed agreements with OpenAI, Google, Microsoft, Amazon Web Services, NVIDIA, SpaceX and Reflection AI to deploy AI on networks.
Anthropic was excluded, labeled a supply-chain risk, and is suing over “all lawful use” clause.
Senator Kirsten Gillibrand questioned Defense Secretary Pete Hegseth on civilian oversight, who affirmed protecting non-combatants.

Blindspot

Center coverage included the detail that Oracle joined the Pentagon’s classified-network AI agreements as the eighth firm, a point not mentioned in Left coverage.
Center coverage also highlighted the stated benefits of the AI tools—streamlining data synthesis, enhancing situational understanding and augmenting warfighter decision-making— which Left coverage did not address.

History

Historical Context

Project Maven began in April 2017 when the Pentagon hired Google to run AI on drone imagery, a $9 million contract that ended after employee protests in 2018. The program’s success spurred a wave of defense AI deals—including the 2014 Third Offset strategy and the 2025 $1.3 billion contract ceiling—leading to the May 1 2026 agreement with eight firms such as Oracle and Amazon.

1 of 4

Oracle Joins Pentagon

Oracle became the eighth AI company to join the Pentagon’s classified-network agreements, expanding the original list of seven firms.
May 1 2026, Department of Defense signed agreements with AI firms to deploy models on Impact Level 6 and 7 networks for lawful use.
Anthropic was excluded after the Pentagon labeled it a supply-chain risk following a safety-guardrails dispute.
Officials said tools will streamline data synthesis, augment decisions, using GenAI.mil accessed by 1.3 million.

1 / 4