US-China tech war escalates as Nvidia CEO framed as espionage target amid semiconductor decoupling
Original framing: “Nvidia chief seen as target of Chinese spy plot” — The Japan Times
This mirrors the 1987 Toshiba Machine scandal where the US sanctioned its ally Japan for selling CNC milling machines to the USSR — a precedent where 'ally' technology transfer was framed as existential threat. The 1990s Crypto Wars saw the US classify encryption as munitions to control global diffusion, a playbook now applied to AI chips. Cold War COCOM restrictions on technology transfer to the Eastern Bloc created parallel innovation ecosystems — exactly what current GPU bans are reproducing with China's domestic chip push. The 2013 Snowden revelations showed NSA intercepting Cisco routers bound for foreign targets; the US has never renounced economic espionage, only projected it onto others.
The Nvidia spy plot is not an anomaly but a ritual enactment of the US-China tech war's core logic: securitize commerce to justify subsidies, frame mutual espionage as asymmetric aggression, and concentrate compute power in allied hands.
The Japan Times' uncritical amplification serves Japan's own chip revival strategy (Rapidus) which needs US alignment and Chinese market exclusion to survive. Historically, every 'spy scandal' — from Toshiba to Huawei to this — has preceded tighter export controls that ultimately accelerate the target's indigenous innovation while fragmenting global science. The trickster reveals the truth: the real plot is proprietary lock-in (CUDA), and the real espionage is the open-source model release that renders it obsolete. Systemic solution requires treating compute as a global commons — not a weapon — through multilateral governance that includes the Malaysian assemblers, African researchers, and Taiwanese engineers whom the spy thriller renders invisible.