Navy bets slow AI rollout is riskier than getting it wrong

The Department of the Navy isn’t just dabbling in AI anymore. A newly signed strategy, greenlit by Acting Secretary Hung Cao, lays out an aggressive blueprint for an “AI-first” fleet, framing sluggish adoption as the real existential threat—not imperfect algorithms. The document, over a year in the making, introduces the “Bits2Effects Cycle,” a five-stage framework designed to weaponize data at machine speed.

The core metric is Mean Time to Effect, or MTTE: the clock ticking between capturing data and triggering a military response. The force that learns fastest wins, the logic goes, and the Navy wants its cycle to be brutally short. By the end of 2026, key infrastructure is supposed to be operational, with a plan to double the ranks of data engineers and AI scientists by 2029. But the philosophical shift is what stands out. In what it calls a “Wartime Approach,” the strategy explicitly states the risks of moving too slowly outweigh the risks of “imperfect alignment.” That’s not a caveat; it’s the central thesis.

Technically, the vision is radical: running large language models and agentic AI directly on warships, even when communications are jammed. Sailors would build their own apps on top of these models, guided by a new “AI War Council” that can pre-approve wartime rule changes. This isn’t theoretical. The Pentagon’s GenAI.mil platform already exploded from 80,000 to 1.5 million daily users in six months. The Navy itself reportedly slashed one submarine planning task from 160 hours to ten minutes using AI.

The human stakes of this acceleration are already visible and politically tangled. During the conflict with Iran, the US military reportedly used Anthropic’s Claude for target analysis. The fallout was swift: the Trump administration cut Anthropic’s access over the company’s insistence on restrictions for autonomous weapons. OpenAI then stepped in, striking a deal to run its models on classified networks while relying on contractual safeguards rather than hard policy lines. The Navy’s new strategy will only intensify this demand signal, pulling the industry deeper into a global AI arms race where China’s military is spinning up procurement requests just as quickly.