Anthropic Accuses Chinese AI Firms of Industrial-Scale Copying as IBM Shares Plunge Over
Photo by Possessed Photography on Unsplash
Anthropic accused DeepSeek and other Chinese AI firms of industrial‑scale copying, saying they used 24,000 fake accounts and 16 million exchanges to distill models, as IBM’s shares fell, Tomshardware reports.
Quick Summary
- •Anthropic accused DeepSeek and other Chinese AI firms of industrial‑scale copying, saying they used 24,000 fake accounts and 16 million exchanges to distill models, as IBM’s shares fell, Tomshardware reports.
- •Key company: Anthropic
- •Also mentioned: DeepSeek, IBM
Anthropic said three Chinese AI labs—DeepSeek, Moonshot and MiniMax—used “industrial‑scale” distillation to copy Claude, deploying 24,000 fake accounts that generated 16 million exchanges, according to Tom’s Hardware. The firm claims the practice violates U.S. export controls and its end‑user license, effectively stealing proprietary model outputs to train smaller rivals.
The allegation arrived as IBM’s stock tumbled sharply after Anthropic announced Claude’s new “Claude Code” capability for automating COBOL modernization, reported by Zerohedge. Bloomberg noted the news spooked IBM investors, who fear the AI tool could erode demand for legacy‑system services that underpin IBM’s mainframe business.
Anthropic’s move underscores a broader U.S. debate over Chinese access to frontier AI models, a theme highlighted by TechCrunch, which linked the accusations to growing concerns about export‑control enforcement and competitive leakage. The company’s statement suggests it will pursue legal and regulatory remedies against the offending firms.
Analysts observing the market reaction note that IBM’s share decline reflects investor anxiety over AI‑driven disruption of its core enterprise software stack, especially as competitors like Anthropic expand into niche code‑generation domains. The combined pressure of export‑control scrutiny and competitive AI advances could reshape revenue streams for legacy‑technology providers.
This article was created using AI technology and reviewed by the SectorHQ editorial team for accuracy and quality.