16th November, 2026
By Anjali Kochhar
A growing underground market is reportedly helping Chinese developers access advanced American artificial intelligence models despite restrictions, with cryptocurrency emerging as a significant payment method.
According to research published by blockchain analytics firm Chainalysis, 40% of 100 surveyed AI proxy service providers accept cryptocurrency payments, primarily through stablecoins. These platforms enable users to access restricted AI models through intermediaries, raising concerns about technology security and regulatory enforcement.
Cryptocurrency Powers the AI Proxy Market
AI proxy services act as intermediaries between users and leading American AI platforms, including ChatGPT and Claude. By routing requests through external servers, these services allow customers to access models that may otherwise be unavailable in their regions.
The investigation found that many providers accept conventional payment methods, including Alipay and WeChat Pay, alongside cryptocurrencies such as USDT.
Stablecoin transactions offer an alternative to traditional banking channels, potentially making cross-border payments easier while complicating efforts to identify the individuals and businesses involved.
China-Linked Crypto Activity Jumps 55%
One of the report’s most significant findings concerns cryptocurrency transactions following major American AI model releases.
Chainalysis observed that payment activity from wallets suspected of having Chinese connections increased 55% above normal levels during the first three days after new frontier AI models became available.
However, approximately 13 days after these releases, activity dropped 21% below baseline levels.
Researchers suggest this pattern could reflect developers seeking early access to advanced models for testing, integration, or knowledge distillation.
Knowledge distillation involves training smaller AI systems using outputs generated by more advanced models, potentially reducing development costs.
Nevertheless, blockchain transaction patterns alone cannot establish how the accessed models were used.
Growing Security and Regulatory Concerns
The findings highlight concerns about the effectiveness of restrictions on advanced AI technology.
The report also references allegations by Anthropic that Chinese AI laboratories used unauthorized accounts and proxy networks to extract capabilities from its Claude models.
Meanwhile, third-party cryptocurrency payment processors reportedly help separate customers from proxy operators, making financial relationships harder to identify.
Chainalysis argues that blockchain intelligence could help investigators trace these payment networks and identify the intermediaries supporting unauthorized access.
Analysis
The findings reveal an emerging intersection between cryptocurrency infrastructure, artificial intelligence competition, and international technology restrictions.
Stablecoins are becoming useful payment tools for digital services operating across borders, including markets facing regulatory barriers.
However, the reported 55% increase in China-linked payment activity demonstrates correlation rather than definitive evidence of unauthorized AI training.
For regulators, the challenge extends beyond restricting access to advanced models. Monitoring intermediary platforms, payment processors, and suspicious financial networks may become equally important.
Ultimately, the report suggests that controlling access to frontier AI will require coordinated technological, financial, and regulatory enforcement rather than relying solely on geographic restrictions.