Anthropic Warn US Government: After OpenAI, Anthropic warn US government on China; say Chinese AI models are stealing our data and the speed … |


After OpenAI, Anthropic warns US government on China; says Chinese AI models are stealing our data and the speed ...

AI giant Anthropic has accused three of China’s largest AI labs — DeepSeek, MiniMax, and Moonshot AI of ‘illicitly’ using its Claude model outputs to train their own systems. According to a report by Business Insider, in a statement Anthropic said that these Chinese albs orchestrated industrial-scale distillation campaigns, creating roughly 24,000 fraudulent Claude accounts that generated more than 16 million exchanges in violation of its terms of service and regional restrictions. For those unaware, distillation is a process where smaller model is trained on the outputs of a larger one, is widely used in AI development. However, Anthropic argues that its Chinese rivals are exploiting the practice to steal capabilities at a fraction of the time and cost it would take to build independently.

Growing concerns across US tech

The warning by Anthropic is followed by the similar concerns raised by OpenAI in January 2025. Last year, ChatGPT-maker OpenAI also alleged that DeepSeek may have improperly used its outputs. Also, earlier this month, Goggle also reported an increase in “distillation attacks.” Anthropic also added that such campaigns are now becoming more sophisticated and intense, urging rapid, coordinated action among industry players, policymakers, and the global AI community.

‘India Well Positioned To Lead The World In AI’: OpenAI CEO Sam Altman At AI Impact Summit

Security risks and export controls

Apart from competitive concerns, Anthropic also warned that improperly distilled models may lack safeguards, potentially enabling dangerous misuse like the development of bioweapon. Anthropic CEO Dario Amodei has been vocal about the need for US export controls, arguing that restricting access to advanced chips could limit both direct model training and the scale of illicit distillation.

Anthropic’s countermeasures and ongoing debate

Anthropic says it has implemented behavioral fingerprinting systems, shares intelligence with other AI companies, and continues to develop new defenses. The company disclosed that during MiniMax’s campaign, its rival pivoted within 24 hours of a new Claude release, redirecting nearly half its traffic to capture fresh capabilities.Representatives for DeepSeek, MiniMax, and Moonshot AI did not respond to Business Insider’s request for comment.



Source link

  • Related Posts

    H-1B Wage Rules: US moves closer to tougher H-1B wage norms after proposal clears federal review

    The United States has moved a step closer to revising wage rules for foreign workers after a proposed regulation affecting the H-1B visa and employment-based green card process cleared federal…

    Bored of Peace? Mike Huckabee’s Israel should ‘take all of Middle East’ remark angers Trump’s Arab allies | World News

    President Donald Trump talks with U.S. ambassador to Israel Mike Huckabee and Israel’s Prime Minister Benjamin Netanyahu as he arrives at Ben Gurion International Airport, Monday, Oct. 13, 2025, near…

    प्रातिक्रिया दे

    आपका ईमेल पता प्रकाशित नहीं किया जाएगा. आवश्यक फ़ील्ड चिह्नित हैं *

    hi_INहिन्दी