Anthropic alleges Alibaba used 25,000 fraudulent accounts to extract Claude data
The Neuron ● Covered by 3 sources
Anthropic accused Alibaba of using 25,000 fraudulent accounts to extract capabilities from Claude through unauthorized API access. The alleged extraction campaign used coordinated accounts, proxy services, and repetitive prompts to harvest reasoning, tool-use, and reinforcement-learning data that could be used to train competing models. The distinction matters because legitimate distillation of intentionally released models should remain permitted, while covert extraction through fraud should face enforcement at the access layer rather than through restrictions on open AI.
Why it matters
Anthropic claims Alibaba used roughly 25,000 fraudulent accounts to collect nearly 29 million Claude interactions, suggesting coordinated industrial-scale extraction rather than casual testing.
Related stories
China closes AI gap with distillation attack
morningbrew.com · 3 months ago ·
33
Anthropic details distillation campaigns from Alibaba, Moonshot AI, and DeepSeek
TechCrunch · 1 month ago ·
32
Likely illegally, Claude gained access to 3 networks. Will Anthropic be held to account?
Ars Technica · 2 months ago ·
25