AI Heist? Anthropic Accuses Alibaba Of Stealing Claude Capabilities Via Fake Accounts

Anthropic alleges Alibaba-linked operators used fake accounts in a massive AI model extraction campaign targeting Claude.

Advertisement
Read Time: 2 mins
Anthropic claimed the alleged activity amounted to an unauthorised AI model distillation campaign.
AI generated image

US artificial intelligence powerhouse Anthropic has levelled serious allegations against Chinese tech major Alibaba Group, accusing it of orchestrating a large-scale operation involving thousands of fake accounts to extract capabilities from its Claude AI models, reported BBC 

The company claimed the alleged activity amounted to an unauthorised AI model distillation campaign. 

According to a letter sent by Anthropic to US Senators Tim Scott and Elizabeth Warren on June 10, operators linked to Alibaba conducted nearly 29 million interactions with Claude through fraudulent accounts, the report said.

Advertisement

The company alleged that the exchanges were aimed at collecting Claude's responses to help develop or improve another AI system without permission.

Anthropic described the alleged operation as one of the largest AI model extraction campaigns it has identified and urged US lawmakers to introduce stronger protections against such practices.

Advertisement

ALSO READ : ChatGPT To Become Faster? OpenAI Unveils Its First AI Chip Jalapeno

The company said the activity involved “distillation attacks”, a technique in which outputs from a more advanced AI model are used to train another model. 

While AI distillation is a recognised development method, Anthropic alleged that the use of fake accounts violated its terms and intellectual property rights.

Advertisement

As reported by the BBC, Alibaba has been contacted for a response to the allegations, while Anthropic has also been asked for further details regarding its claims.

Anthropic argued that such unauthorised extraction could allow competitors to replicate years of AI research at significantly lower costs and potentially impact billions of dollars invested in AI development.

The allegations come amid growing competition between US and Chinese AI companies over advanced models, data, and intellectual property. Similar concerns over AI model copying have previously been raised by other firms, including OpenAI. 

Anthropic has called for regulatory action to deter such activities and strengthen protections for advanced AI systems.

ALSO READ : Amazon Pledges $48 Billion India Investment To Scale AI Infra By 2030: CEO Andy Jassy Meets PM Modi

Advertisement

Essential Business Intelligence, Sharp Market Insights, Practical Personal Finance Advice, Daily Fuel, Gold and Silver Prices and Latest Stories — On NDTV Profit.

Loading...