GMAsia
    🇸🇬Singapore·AI News·9 Sept 2026·via Malay Mail

    Washington slams Chinese AI firms for ‘systematic’ theft of US model capabilities

    Washington has accused several prominent Chinese artificial intelligence laboratories, including DeepSeek and Moonshot AI, of systematically extracting capabilities from American AI models. This accusation, detailed in an advisory from the US cyber defense agency, comes as the two nations intensify their competition for technological supremacy. The advisory claims Chinese firms are using industrial-scale knowledge distillation campaigns, a method where an AI model mimics a larger one, to illicitly copy US systems from companies like OpenAI, Anthropic, Google, and xAI. These actions, which allegedly began in late 2024 and likely have Chinese government awareness, allow Chinese companies to significantly shorten development timelines and reduce costs for frontier AI models. The US government has previously made similar accusations of AI theft through distillation.

    Nexa's Summary

    The US accusation against Chinese AI firms like DeepSeek and Moonshot AI for systematic theft of American model capabilities through distillation points to a deepening tech rivalry. While distillation is a common industry practice, Washington claims Chinese companies are deploying it at an industrial scale to gain an unfair advantage, reducing their AI development timelines and financial expenditures. This directly impacts the competitive landscape for Asian AI startups and established tech giants. For example, Alibaba, also named in the advisory, faces increased scrutiny on its AI development methods, which could affect its international partnerships and market access. The timing of this advisory, ahead of a potential meeting between President Trump and President Xi Jinping, suggests AI will be a key point of contention. The US is under pressure to respond to the growing success of Chinese AI models, which are often preferred for their lower cost compared to more powerful but pricier US versions. This dynamic creates a complex environment for Asian markets, where companies must navigate both the allure of cost-effective Chinese AI and the risks associated with US regulatory actions. The Chinese foreign ministry's denial of unfounded accusations underscores the diplomatic tension surrounding AI development. The thing to watch is how this impacts the global supply chain for AI training data and model access. If the US implements stricter controls or bans on foreign-made AI models, as some startups have urged against for non-government use, it could fragment the global AI ecosystem. This would force Asian companies to choose between US and Chinese technologies, potentially hindering innovation and increasing costs across the region.

    #washington#deepseek#openai#donald trump#xi jinping#moonshot ai
    Original reporting by Malay MailWe don't republish, read the full story →

    Related reading

    6 stories