U.S. Invokes Section 232 to Levy 25 % Duty on Nvidia H200 and AMD MI325X AI Chips

Washington last invoked Section 232 this aggressively in 2018 for steel and aluminium; before that, the 1986 U.S.–Japan semiconductor accord similarly used trade pressure to on-shore chip capacity. Today’s move fits a decades-long arc of techno-nationalism and supply-chain securitisation accelerated by the 2020–21 pandemic, CHIPS Act 2022, and rising U.S.–China strategic rivalry. By taxing rather than banning exports, the administration experiments with a revenue-sharing model reminiscent of Britain’s Navigation Acts (1651) that taxed colonial trade to fund domestic shipbuilding. If the tariff widens, it could redraw manufacturing geographies much like the 1930 Smoot-Hawley duties shifted global agricultural flows—yet it may also simply nudge firms to re-route, delaying but not reversing Asia’s fabrication dominance. Whether this is a pivotal inflection or a footnote will hinge on follow-on tariffs and the still-uncertain build-out of U.S. fabs over the next decade; on a century scale, it illustrates how great powers periodically weaponise chokepoint technologies when industrial leadership feels at risk.