EXPLORE THIS STORY
DEEPSEEK V4 LAUNCHES WITH 1.6 TRILLION PARAMETERS AND HUAWEI'S BACKING: THE AI WAR TILTS
AI-generated content — Analyses are produced by artificial intelligence from press articles. They may contain errors or biases. Learn more
Washington responds to DeepSeek V4 with an anti-espionage offensive: Kratsios memo accuses China of industrial theft
Dominant angle identified — does not reflect unanimity of this country’s media
Washington doesn't look at DeepSeek V4 through a technologist's eyes -- it looks through a prosecutor's. The same day, NPR reveals a memo from Michael Kratsios, the president's science adviser, accusing foreign entities 'principally based in China' of conducting 'deliberate, industrial-scale campaigns' to 'distill' -- extract capabilities from -- leading US AI models. The administration pledges to work with American companies to identify activities, build defenses, and punish offenders.
The memo arrives as the US-China AI gap has 'effectively closed,' according to a Stanford report. China's embassy in Washington replied that Beijing 'has always been committed to promoting scientific and technological progress through cooperation and healthy competition' and 'attaches great importance to intellectual property protection.'
The House Foreign Affairs Committee has unanimously backed a bill to sanction foreign actors extracting 'key technical features' from closed-source US AI models. Rep. Bill Huizenga calls these extractions 'the latest frontier of Chinese economic coercion.' The timing is not coincidental: the Kratsios memo drops the day DeepSeek announces V4. Washington could not let the launch pass without showing it's fighting back.
Security framing transforms Chinese innovation into threat -- no analysis of V4's technical merits
Memo timing on V4 launch day suggests political messaging over strategic response
Theft rhetoric ignores that DeepSeek V4 is open-source -- is open-source theft?
Discover how another country covers this same story.