OpenAI signs new deal, to bring its AI to US military through Amazon

OpenAI has signed a new agreement to sell access to its AI models to US defence and government agencies through Amazon Web Services (AWS), covering both classified and unclassified operations. The deal marks an escalation in the ChatGPT-maker’s ambitions in the government sector even as its rival Anthropic has been declared supply chain risk.Anthropic, whose Claude models have been used multiple times by the Pentagon, worked alongside Palantir and AWS to deploy its models in classified military and intelligence systems. In July 2025, it won a Pentagon contract worth up to $200 million.But that relationship developed cracks in February when Anthropic refused to allow unrestricted military use of its AI in use cases involving domestic surveillance and autonomous weapons. The Pentagon labelled Anthropic a “supply chain risk”, effectively shutting it out of government work entirely. That left a gap. OpenAI moved quickly to fill it.

What OpenAI’s deal involves

According to Reuters, OpenAI had previously limited its government work to unclassified use cases but now, the company has secured a Pentagon contract that extends its reach into classified operations, and its new partnership with AWS is the infrastructure backbone that makes it possible.Notably, AWS is already deeply embedded in US federal systems, making it a natural channel for OpenAI to reach defence and intelligence customers at scale.Moreover, Amazon and OpenAI announced partnership last month wherein the cloud provides will co-create a Stateful Runtime Environment powered by OpenAI models, available on Amazon Bedrock for AWS customers to build generative AI applications and agents at production scale. Other key highlights of the last month’s deal include: