The headline was simple and unglamorous: Breitbart hosted a policy event called “AI Is Here: What to Fear, What to Embrace.” The real news was the mood on stage — a Department of War that is sprinting to adopt generative AI, led by Under Secretary of War for Research and Engineering Emil Michael, and an industry eager to sell tools that promise to reshape government work. This wasn’t a theoretical panel. It was a white-hot push to move AI out of labs and into everyday government use, with all the promise and peril that brings.
Emil Michael: Making AI the Department’s No. 1 Priority
Under Secretary Emil Michael made no effort to downplay the scale of the effort. He’s calling applied AI the number one technology priority at the Department of War, and the department’s GenAI.mil platform is already rolling out to large parts of the workforce. That’s the right direction: America should be first to adopt powerful tools that boost our military edge and government efficiency. But “move fast” without clear, public guardrails is how you get embarrassing data spills or messy vendor entanglements.
Procurement, Security, and the Anthropic Shadow
Onstage optimism collided with a harder reality: the Anthropic supply‑chain designation and the resulting lawsuit have exposed real tensions in the Department’s vendor strategy. The government can and must vet vendors for national security reasons. Yet if procurement moves are opaque, they’ll chill private innovation and hand competitors a playbook for legal fights. The department needs a firm, transparent vetting process — not secret lists and courtroom surprises that look like bureaucrats playing whack‑a‑mole.
Balancing Guardrails and Growth
GenAI.mil shows the department wants the benefits: faster reports, smarter decision support, and scaled expertise. Industry panelists rightly pointed to business efficiencies and the way tools can empower ordinary workers — no need for a Yale degree to get results. Still, the public deserves to know the guardrails. How are frontier models validated? Who approves classified use? What are the conflict‑of‑interest safeguards for officials picking vendors? Those aren’t “nice to haves.” They’re essential.
Why Conservatives Should Care — And Stay Loud
Conservatives should welcome a Department that embraces American tech and muscle. But we should also demand accountability. Support for military advantage doesn’t mean blind faith in procurement bureaus. If the administration and Emil Michael want to accelerate AI in government, give conservatives two things: clarity and results. Release the GenAI.mil governance playbook. Publish adoption metrics and real use cases. Let the public see how security and liberty are being protected.
Bottom line: this Breitbart event made clear that AI is no longer a promise — it’s a government program in motion. That’s good news for American strength, if handled properly. It’s a problem if handled quietly. So applaud the ambition, but don’t stop asking the inconvenient questions. The stakes are too high for cheerleading without oversight.

