OpenAI Reaches Pentagon Deal Hours After Anthropic Ban

OpenAI has reached a deal with the Pentagon, announced within hours of the Trump administration ordering the federal government to stop using Anthropic, according to multiple news reports.

The agreement makes OpenAI the latest major artificial intelligence company to secure a foothold inside the Defense Department as federal agencies reshuffle which commercial AI tools they can use. The timing underscores how quickly the government’s AI vendor landscape can change following an administration directive.

Details of the deal were reported by outlets including CNN, NBC News, NPR and Semafor. Those reports described the arrangement as a Pentagon agreement involving OpenAI and referenced safeguards tied to the work.

The Trump administration’s move to ban Anthropic from government use was reported by CNN, NBC News and NPR. The reports framed the decision as an order directing federal agencies to halt use of Anthropic, which immediately raised questions across government and industry about continuity for teams that had been using the company’s technology.

Taken together, the two developments signal an abrupt shift in which companies can provide AI capabilities to national security customers. The Pentagon and other agencies have been accelerating their adoption of AI tools for analysis, workflow support and other tasks, and vendor access can determine how quickly programs move and which models become embedded across systems.

The deal also matters because it places OpenAI more directly into the defense contracting environment, a space that typically comes with heightened scrutiny, requirements around data handling, and oversight expectations. Reports emphasized safeguards, reflecting the sensitivity of deploying advanced AI systems in defense contexts and the need to set boundaries around use, security and accountability.

The administration’s decision to blacklist Anthropic could have ripple effects beyond the Pentagon. When a vendor is barred from federal use, agencies may need to pause pilots, revisit procurement plans and transition work to other platforms, potentially affecting internal operations and timelines.

What happens next will depend on how quickly federal agencies implement the Anthropic ban and how the Pentagon operationalizes its new work with OpenAI. Agencies and contractors that relied on Anthropic’s tools may be required to identify alternatives, update contracts and shift technical integrations, while OpenAI’s Pentagon deal will move into implementation under the terms described in the agreement.

The Pentagon and the administration are expected to face continued questions about the scope of the Anthropic prohibition, the standards applied in making the decision, and how safeguards will be enforced as OpenAI’s work proceeds.

For the AI industry, the message is immediate: in Washington’s rapidly evolving national security AI push, access can be granted or revoked in a matter of hours, reshaping the competitive landscape overnight.