Pentagon signs AI agreements with seven companies, leaving out Anthropic

Washington: The Pentagon said on Friday it had reached agreements with seven artificial intelligence companies to bring their advanced tools onto the Department of Defense’s classified networks, as it moves to widen the pool of AI providers working with the US military.

In a statement, the Pentagon said SpaceX, OpenAI, Google, NVIDIA, Reflection, Microsoft and Amazon Web Services would be integrated into its Impact Levels 6 and 7 network environments. The move will expand access to their products across the military.

The announcement did not include Anthropic, which has been in dispute with the Pentagon over safeguards governing military use of its AI systems. Earlier this year, the Pentagon designated the company a supply-chain risk and prohibited its use by the department and its contractors, despite the company’s tools being widely used across the military.

Push to broaden AI access

The Pentagon said the expansion of AI services for troops is intended to reduce dependence on any single provider. In its statement, it said the broader offering would help avoid

vendor lock

.

According to the Pentagon, military personnel use AI for planning, logistics, targeting and other functions aimed at making large-scale operations more efficient and faster.

The department also said its main AI platform, GenAI.mil, has been used by more than 1.3 million Department of Defense personnel in the five months since it began operating.

Several of the companies included in the new agreements already have ties with the Pentagon. Earlier this week, a source told Reuters that Google had signed a separate agreement allowing the Department of Defense to use its AI models for classified work.

Anthropic dispute remains unresolved

Reuters reported that Pentagon staff, former officials and IT contractors working closely with the US military were reluctant to stop using Anthropic’s tools, which they regarded as better than alternatives, even after being instructed to phase them out over the next six months.

On Friday, Department of Defense Chief Technology Officer Emil Michael told CNBC that Anthropic remained a supply-chain risk. He also said that Mythos, the company’s AI model with advanced cyber capabilities that drew concern among US officials and corporate America because of its potential to boost hackers, represented a separate national security issue.

While a number of companies and public and private organisations have been given access to a Mythos preview product to help protect their IT systems against future cyberattacks, it remains unclear whether the Pentagon is part of that programme.

US President Donald Trump said last week that Anthropic was “shaping up” in the view of his administration, suggesting the company could still reverse its blacklisting by the Pentagon.

The dispute, however, has underscored the Pentagon’s effort to diversify its AI supply base, a shift that could create openings for smaller defence-focused AI startups.