China’s Alibaba Bans Anthropic AI for Employees After ‘Distillation Attack’ Accusation

China’s e-commerce giant Alibaba has banned its employees from using Anthropic’s AI tools, including Claude Code, following accusations of a “distillation attack.” The ban is part of Alibaba’s efforts to strengthen its cybersecurity posture and protect sensitive data.

According to sources, Alibaba has put Claude on a high-risk software list, effectively preventing employees from using it for work purposes. This decision is likely driven by concerns over data security and intellectual property protection. Alibaba’s own AI assistant, Qoder, will be the preferred choice for employees going forward.

The accusations against Alibaba are serious. Anthropic claims to have been subjected to a “distillation attack,” which involved extracting its AI capabilities illicitly. This incident reportedly represents the largest known distillation attack on Anthropic to date. In response, Alibaba has accused Anthropic of attempting to extract its own AI data through back doors.

The implications of this saga extend beyond the specifics of the case itself. It highlights the increasing tensions between China and the West over technology security and intellectual property protection. This is not the first time Alibaba has faced criticism over its handling of sensitive data; in 2020, the company was fined $2.8 billion for antitrust practices in China.

The situation also raises questions about the use of artificial intelligence tools by companies with significant interests in both China and the West. Anthropic’s terms of service explicitly prohibit Chinese companies from using its models, citing concerns over security risks. However, it appears that some companies have found ways to circumvent these restrictions, leading to accusations of back-door access.

Anthropic has taken steps to address these concerns by closing loopholes that allowed Chinese companies to bypass its restrictions. According to a report by The Financial Times, the company has identified instances where Chinese companies used third countries as proxies to access its AI tools. This move may indicate a willingness by Anthropic to strengthen its security posture and protect user data.

However, this development also raises questions about the effectiveness of such measures in preventing data theft or intellectual property misuse. Can companies like Alibaba really trust that their employees will not attempt to circumvent restrictions on using sensitive AI tools? The answer remains unclear.

The use of AI tools by companies with significant interests in both China and the West highlights several key issues. One is the role of technology firms in addressing security concerns and protecting intellectual property. Another is the delicate balance between cooperation and competition in the global tech landscape. Finally, there is the issue of regulatory frameworks that govern the use of AI tools across borders.

Alibaba’s decision to ban Anthropic’s AI tools may be seen as a shot across the bow for companies seeking to exploit back doors and circumvent security restrictions. However, it also raises questions about the effectiveness of such measures and highlights the need for greater cooperation between technology firms, governments, and regulatory bodies to address the complex challenges posed by the use of AI tools in global business.