The ReadJune 24, 2026via Financial Times Technology
Anthropic accuses Alibaba of obtaining ‘illicit’ access to Claude
Why it matters
As AI models become core IP, companies face new security threats from competitors using account fraud and API scraping to extract capabilities. This case signals a shift from traditional industrial espionage to model extraction attacks—and raises urgent questions about API access controls and international enforcement.
Key signals
- Anthropic alleges Alibaba used fake accounts to access Claude API
- Accusation centers on 'extracting' model capabilities
- Involves Chinese ecommerce giant and US AI lab—geopolitical/trade angle
- First public case of alleged model IP theft at scale
- Raises questions about API security, account verification, and ToS enforcement
The hook
Anthropic accuses Alibaba of reverse-engineering Claude through fake accounts. The first major IP theft case in the AI arms race.
AI company says the Chinese ecommerce giant used fake accounts to ‘extract’ chatbot’s capabilities