When 'Chinese AI Could Counter U.S. Military Power' Reads More Like Aspiration Than Capability

This report references a CNAS analysis warning that Chinese AI could be used to counter U.S. military operations. We attribute the claim to CNAS and treat it as an assessment, not as proven battlefield capability.

The substance is familiar terrain for us: AI is a dual-use technology, and China has been closing the gap with the Western frontier at speed, particularly in open-weight models. That progress is real. But military-framed warnings have a structural tendency to merge three different things — what a system can do in a lab, what it could do if perfected, and what an adversary might intend — into a single alarming sentence. The phrase 'could counter U.S. military operations' is doing a lot of work, and the gap between 'could' and 'does' is exactly where sober analysis lives.

Why it matters: reports like this shape budgets, export controls and the political temperature of the U.S.–China rivalry. And as we've argued before, aggressive export restrictions can cut both ways — they can slow an adversary, or they can accelerate the push toward self-sufficiency. Regulating the panic rather than the demonstrated capability is the recurring failure mode.

Our reading: take the warning seriously as a prompt to govern dual-use AI well, not as a scoreboard verdict on who 'wins' a future conflict. The strategically important question is not whether Chinese AI is catching up — it is — but who governs how this technology is deployed, on both sides. The optimistic long-term case doesn't deny great-power competition; it insists that the same capabilities feeding military anxiety are the ones that, governed wisely, accelerate medicine, energy and science. The task is to keep the dual-use conversation grounded in evidence rather than in the most quotable worst case.