Anthropic, a US artificial intelligence lab, has alleged that Chinese AI developers used fake accounts and proxy networks—intermediaries that mask identity—to access and extract data from Claude, its flagship AI model. US security agencies including the National Security Agency and FBI issued an advisory last month stating that Chinese AI developers have been systematically extracting capabilities from US frontier models—advanced AI systems at the cutting edge of the field—to reduce their own computing and electricity costs.
In some cases, Anthropic claims Chinese firms even used Claude itself as a hidden backend—the behind-the-scenes infrastructure—for their own applications. US Treasury Secretary Scott Bessent warned in July that Chinese companies were engaged in large-scale 'distillation', a technique that compresses knowledge from one model into another.
Why it matters
- Schools and parents should understand that AI security and access restrictions affect which tools students can use globally.
- The allegations highlight geopolitical tensions around AI development that may shape future regulations for Hong Kong and Singapore.
- If true, the claims suggest companies must protect user data more carefully when accessing popular AI platforms.
What to do
- Stay informed about AI policy developments affecting educational platforms your school or child uses.
- Review terms of service when using AI tools to understand data security and account protection.
- Follow announcements from Singapore's Infocomm Media Development Authority or Hong Kong authorities on AI governance.