SAN FRANCISCO, Feb. 28, 2026 (GLOBE NEWSWIRE) — Following the Trump Administration’s decision to exclude Anthropic from federal use and the Pentagon’s reported deal with OpenAI, an executive branch overseeing active defense deployments is available to assess the impact on procurement and governance.
Ben Van Roo, CEO of Legion Intelligence, leads a company that deploys agentic AI systems in operational environments within the U.S. Departments of War and Energy, and argues that the current public framing of the dispute obscures the operational realities within those environments.
“In practice, these models are not used for autonomous lethal actions. They are used to accelerate human analysis, reveal intelligence gaps and reduce decision latency. The public debate is on autonomous weapons. The operational reality is analytical acceleration.”
He has publicly challenged the emerging dynamic in which private AI labs enforce product-level restrictions that effectively determine which legal use cases are allowed by the government.
“Once embedded in operational systems, base models behave less like software subscriptions and more like infrastructure dependencies. If this impasse continues, procurement will not stagnate. It will be reallocated within the quarter. Demand within the national security system does not disappear because a policy team becomes uncomfortable.”
With active US and allied strikes on Iranian targets, analytical and operational timelines have been further compressed. “In real crises, capability gaps are not theoretical. They change outcomes.” said Van Roo.
Van Roo can focus on:
- How LLMs are currently implemented within classified analytical and operational workflows
- National security risks are on the table as the Pentagon moves from Anthropic to OpenAI amid conflict with Iran
- What ‘legal use’ means in the context of procurement and missions
- The practical effect of product-level divestitures during active crises
- How adversarial model distillation and open source replication complicate unilateral mitigation strategies
- The structural gap between corporate AI policy decisions and democratic oversight mechanisms
Available for on-the-record interviews, background briefings and rapid response.
Media contact:
Kyle Arteaga
[email protected]
