Pentagon embraces AI risk over slow adoption in Navy strategy

The Navy's new strategy centers on a five-stage framework called the 'Bits2Effects Cycle,' which traces military data from automated collection through transmission, classification, and analysis to its use in real tactical decisions and actions. The framework feeds lessons learned back into the cycle to allow continuous updates to systems, tactics, and training. The key performance metric is 'Mean Time to Effect' (MTTE)—the window from when new data is captured until it produces a concrete military response or adaptation. According to the strategy paper, in a drawn-out conflict with multiple learning cycles, the force that learns and adapts fastest will dominate. Deputy Secretary of the Navy Cao said the strategy would let the Department of the Navy 'out-learn and out-fight any adversary' through rapid deployment of data and AI, describing it as a roadmap for building an 'AI-first' fleet that turns information into military advantage and enables faster, better decision-making.

The strategy lays out six explicit goals: speed up operational AI deployment, improve data availability and usability, expand technical infrastructure, streamline approval processes, strengthen data and AI literacy among personnel, and deepen collaboration with industry, academia, government agencies, and allies. Major measures are supposed to be in place by the first quarter of fiscal year 2027 (ending December 2026). By the end of fiscal year 2029, the number of qualified data engineers, data scientists, and AI and machine learning engineers is supposed to double. A particularly significant element is the plan to run large language models and agentic AI directly on warships and with Marine Corps expeditionary units, with these systems designed to work even when communications are jammed or cut off. Service members would build their own applications on top of them. An 'AI War Council' would prioritize use cases, coordinate resources, and pre-approve wartime changes to data sharing, classification, and deployment rules.

The strategy makes an explicit risk trade-off that sits at the center of broader Pentagon AI policy: the risks of moving too slowly outweigh the risks of 'imperfect alignment' in these systems. This passage is framed within a 'Wartime Approach' context, in which the Department of Defense wants to handle risk assessments and organizational hurdles as if the country were already at war, making decisions that favor speed. This philosophy reflects observable progress in military AI already in use. GenAI.mil, the central Defense Department platform where personnel can use generative AI, hit 1.5 million daily users in June 2026, up from 80,000 when it launched in December 2025. Uses range from routine office tasks to military planning and combat operations. The Army is testing AI in a 'Next Generation Command and Control' system to process large volumes of data faster and help soldiers build situational awareness and make faster decisions. A Navy AI program reportedly cut a submarine planning task from 160 hours down to ten minutes.

The urgency is driven partly by global competition. China is pushing military AI adoption at a rapid pace. Researchers at Georgetown University analyzed thousands of publicly available procurement requests from the Chinese People's Liberation Army and found Beijing testing AI systems for unmanned combat vehicles, cyber defense, ship tracking, target acquisition on land, at sea, and in space, and deepfake-powered disinformation. NATO is already using AI operationally as well—French Admiral Pierre Vandier, NATO's top officer for digital transformation, said alliance members are using AI to track Russia's shadow tanker fleet. Israel deployed AI to sift through intercepted intelligence data ahead of its war against Iran. The U.S. military is investing heavily in integrating commercial AI and plans to let AI companies train military-specific model versions on classified data, which would bake sensitive intelligence directly into the models. The U.S. military has already deployed Anthropic's Claude for target analysis and strike planning during conflict with Iran—a politically charged deployment because the Trump administration locked Anthropic out of government systems after the company insisted on restrictions for fully autonomous weapons and mass domestic surveillance. Shortly after, OpenAI struck a deal with the Pentagon to run its models on classified networks, citing similar red lines but relying on contractual and technical safeguards rather than hard policy demands. The Navy's strategy is likely to push military demand for powerful language models and AI agents even higher.

Cybersecurity is where the stakes are described as highest. Zhou Hongyi, founder of Chinese cybersecurity firm Qihoo 360, drew an explicit parallel to nuclear escalation, arguing that the ability of AI models like Anthropic's Claude Mythos to autonomously find vulnerabilities and build attack chains amounts to 'cyber nuclear weapons of the AI age.' The UK's AI Security Institute revised its estimate for how fast AI cyber capabilities are doubling, adjusting it upward twice in just a few months. The U.S. government now treats these models as strategic assets and initially blocked Anthropic from publicly launching its Fable 5 AI model. Zhou called Fable 5 a 'civilian, neutered version of Mythos' and suggested the U.S. feared foreign actors would jailbreak the system to reach Mythos-level capabilities, stating: 'This is what the US government finds most intolerable. It must ensure that it alone possesses this capability, forming an absolute monopoly over this strategic asset.' The European Union, meanwhile, is stuck on the sidelines, dependent on the goodwill of big U.S. tech companies because comparable European products don't exist.

The Pentagon's Navy strategy reflects a fundamental shift in how the U.S. military views artificial intelligence risk. Rather than treating safety and alignment concerns as the primary constraint on AI deployment, the strategy reframes slow adoption as the bigger strategic danger. This is not a decision made in isolation: it sits within a broader Department of Defense AI strategy and appears driven by the observed reality that adversaries are moving faster. China, according to analysis of procurement documents by Georgetown University researchers, is testing AI systems for unmanned combat vehicles, cyber defense, ship tracking, and target acquisition across land, sea, and space. NATO allies are already using AI operationally—French and Israeli forces have deployed AI for intelligence analysis and tracking Russian assets. For the U.S. military, falling behind in AI deployment speed poses a concrete tactical risk.

The practical implications are already visible. The Navy has run large language models directly on warships and with Marine Corps units, designed to function even when communications are jammed. A Navy AI program reportedly reduced a submarine planning task from 160 hours to ten minutes. The U.S. military has deployed Anthropic's Claude for target analysis and strike planning during recent conflicts. GenAI.mil's growth from 80,000 to 1.5 million daily users in seven months shows that adoption is not theoretical—it is happening at scale. This creates enormous demand pressure on AI companies: OpenAI recently won a Pentagon contract to run models on classified networks, and the Navy strategy will likely drive that demand higher.

Cybersecurity represents the highest-stakes domain. The UK's AI Security Institute has adjusted its estimates for how fast AI cyber capabilities are doubling upward twice in recent months, signaling a measurable acceleration in technical progress. Chinese cybersecurity researcher Zhou Hongyi has explicitly compared autonomous AI vulnerability-finding and attack-chain building to 'cyber nuclear weapons of the AI age.' The U.S. government initially blocked Anthropic from publicly launching its Fable 5 model, reportedly out of concern that foreign actors could jailbreak it to access the underlying Mythos capabilities. This reflects a view of frontier AI as a strategic asset that must be protected through monopoly control. Meanwhile, the European Union lacks comparable AI capabilities and is dependent on the goodwill of U.S. tech companies.

No comments yet. Be the first to share your thoughts!