Key Takeaways
- US agencies accuse six Chinese AI firms of industrial-scale attacks.
- The firms are alleged to have extracted capabilities from US models.
- The actions are seen as reducing development timelines and costs.
The United States has accused six Chinese AI firms of engaging in industrial-scale attacks to copy advanced US AI models, according to a joint statement from the National Security Agency (NSA), Cybersecurity and Infrastructure Security Agency (CISA), and Federal Bureau of Investigation (FBI).
The accused firms, DeepSeek, Moonshot AI, Alibaba, MiniMax, StepFun, and Z.AI, are alleged to have extracted capabilities from US models such as Claude, GPT, Gemini, and Grok, since at least late 2024.
The agencies stated that these firms 'likely' acted with 'Chinese government awareness' when extracting the capabilities, suggesting a state-sponsored or state-endorsed involvement.
The actions are described as significantly reducing the AI development timelines and financial expenditures for the Chinese firms, potentially sparing billions in development costs.
The US agencies emphasized that China-based AI companies that conduct such attacks see shorter development timelines and reduced financial expenditures, highlighting the competitive advantage gained through these practices.
The statement from the US agencies comes amid growing concerns over the rapid advancement of AI technology and the potential for such practices to undermine global innovation and security.
The accused firms have not commented on the allegations, and their representatives could not be reached for comment.
The US agencies stressed that the actions are not only illegal but also pose a significant threat to the global AI ecosystem, potentially leading to a race to the bottom in terms of ethical and legal standards.
