Chinese AI companies 'distilled' Claude to improve own models, Anthropic says - Reuters
Chinese AI companies 'distilled' Claude to improve own models, Anthropic says Reuters
Coverage by Political Leaning
See how different sides of the spectrum reported this story
Notable Quotes
""These campaigns are growing in intensity and sophistication. The window to act is narrow, and the threat extends beyond any single company or region.""
— Anthropic , Executive
"The scary part is, you can take all of the power and unleash it, because you don’t have anyone that actually enforces those guardrails on the other side."
— Gal Elbaz , Executive
"We have high confidence these labs were conducting distillation attacks at scale."
— Jacob Klein , Executive
"What we can say with confidence is they distilled us at scale."
— Jacob Klein , Executive
"These campaigns are growing in intensity and sophistication."
— Claude , Other
"We have high confidence these labs were conducting distillation attacks at scale."
— Jacob Klein , Executive
Key People
Anthropic is a technology company focused on AI research and development.
Claude is an AI model developed by Anthropic.
DeepSeek is a Chinese AI firm targeting major AI companies.
Gal Elbaz is the co-founder and chief technology officer of Oligo Security.
Jacob Klein is the head of threat intelligence at Anthropic.
Kyrsten Sinema is a U.S. Senator from Arizona.
MiniMax is a Chinese AI firm focused on model training.
Moonshot is a Chinese AI company involved in AI model development.
The President of Venezuela, involved in a long-standing political crisis.
Tags
All Coverage
<a href="https://news.google.com/rss/articles/CBMisAFBVV95cUxNMWN4eF9ybjZaS0NxNE8zZDFZUjd2R3dwVFZERXlCQjZhb2prRGhLTXJRaUU3RV81TEpXM29yUGJPQ1F2Skl2eGo2T2s5aEVYZUFfc0lhMy1FZ1hVMWVXdWY3dEk4MWNWTzhHa3Y2LS1oTTVmSHowdjJaTTV2dGU3MGE1YmxldXJsOFRNamtFS21fZkdMTFoySlByVWRSaGltOTRORk1pYkpqd2dZVWJ1Sw?oc=5" target="_blank">Chinese AI companies 'distilled' Claude to improve own models, Anthropic says</a> <font color="#6f6f6f">Reuters</font>
Anthropic alleges that three Chinese AI laboratories—DeepSeek, Moonshot AI, and MiniMax—used approximately 24,000 fraudulent accounts to generate over 16 million exchanges with its Claude chatbot in a coordinated 'distillation' campaign.
Anthropic reports that three Chinese AI companies—DeepSeek, Moonshot, and MiniMax—used 24,000 fake accounts to generate over 16 million interactions with its Claude model, employing 'distillation' techniques to illicitly enhance their own models.
Anthropic alleges that three Chinese companies used Claude to improperly obtain capabilities to enhance their own models, violating terms of service and regional access restrictions.
Anthropic accuses three Chinese AI companies of using Claude to improperly obtain capabilities to improve their own models, violating terms of service and regional access restrictions.
Anthropic alleges that DeepSeek, Moonshot AI, and MiniMax used approximately 24,000 fraudulent accounts to generate over 16 million exchanges with Claude in a coordinated 'distillation' campaign.
Anthropic reports that three Chinese AI firms used 'distillation' techniques to illicitly enhance their models by interacting with Claude through 24,000 fake accounts, violating terms of service and regional access restrictions.
Anthropic alleges that three Chinese AI companies—DeepSeek, Moonshot, and MiniMax—used approximately 24,000 fake accounts to generate over 16 million interactions with its Claude model, violating terms of service and regional access restrictions.
Anthropic accuses three Chinese AI laboratories—DeepSeek, Moonshot, and MiniMax—of using 'distillation' techniques to extract capabilities from its Claude model, potentially posing national security risks.
Similar Stories
Related coverage based on topic and tags
China's Tencent releases new open-source AI model for coding, research tasks - Reuters
China's Tencent releases new open-source AI model for coding, research tasks Reuters
August 28, 2026 at 09:37 AMAnthropic unveils new framework allowing AI agents to operate physical devices - Reuters
Anthropic unveils new framework allowing AI agents to operate physical devices Reuters
August 27, 2026 at 06:49 PMSalesforce raises annual forecasts, expands AI partnership with Anthropic - Reuters
Salesforce raises annual forecasts, expands AI partnership with Anthropic Reuters
August 26, 2026 at 11:50 PMAlibaba launches Wan3.0 AI video model after $10 billion share sale - Reuters
Alibaba launches Wan3.0 AI video model after $10 billion share sale Reuters
August 24, 2026 at 08:23 AMAnthropic to rent AI computing power from Nscale for $45 billion, source says - Reuters
Anthropic to rent AI computing power from Nscale for $45 billion, source says Reuters
August 26, 2026 at 06:35 PMChina’s humanoid robots aren’t smart enough to take your job – yet - Reuters
China’s humanoid robots aren’t smart enough to take your job – yet Reuters
August 27, 2026 at 05:00 AM