Anthropic Accuses Alibaba of Massive Claude AI Theft Attempt

Anthropic alleges that Alibaba used nearly 25,000 fake accounts to run 28.8 million queries in an attempt to replicate its Claude AI model

Justin Tomlinson

Editor-in-Chief, Mora Discover

2 sources
Anthropic Accuses Alibaba of Massive Claude AI Theft Attempt

Artificial intelligence startup Anthropic has accused its Chinese rival Alibaba and its Qwen AI unit of attempting to steal its technology through a massive data extraction campaign. Anthropic alleges that Alibaba engaged in the largest-known AI distillation campaign to date, which was aimed at replicating the capabilities of its Claude AI model.[1][2]

According to the allegations, Alibaba utilized nearly 25,000 fake and fraudulent accounts to access the Claude system. These accounts were reportedly used to generate 28.8 million queries and interactions, allowing the Chinese firm to illicitly extract data and replicate Anthropic's proprietary AI technology.[1][2]

Related stories

Exclusive: Inside Amazon’s brutal AI-centric app-ification of HR
Exclusive: Inside Amazon’s brutal AI-centric app-ification of HR
Fast Company

Exclusive: Inside Amazon’s brutal AI-centric app-ification of HR

At an Amazon warehouse in Northeast Ohio, Laura works the night shift, operating an order picker that puts her 40 feet in the air. “I’m a treasure hunter,” she says. “I use a forklift to go find stuff.” The warehouse has not yet adopted any of the cutting-edge robotics systems that are transforming other Amazon facilities —but automation is creeping in. In the 18 months since she started this job, Laura—who asked to use a pseudonym out of fear of retaliation—has seen a major change in the HR presence at her warehouse. At most Amazon warehouses, a team of HR employees used to be available on site to help workers resolve issues as they arise, reachable for most of their shift. But something changed this year. “Now it’s like HR is there on banking hours,” Laura says, noting that she never see