Alibaba has moved to ban its employees from using Anthropic’s artificial intelligence tools for work purposes, effective July 10. This significant decision stems from concerns over potential “back-door” security risks associated with the U.S. company’s technology. The e-commerce giant has reportedly placed Anthropic’s Claude code on a high-risk software list, according to individuals familiar with the internal operations who spoke on condition of anonymity.
This development follows a public dispute between the two companies in June. Anthropic had sent a letter to the U.S. Senate Committee on Banking, Housing, and Urban Affairs, accusing Alibaba of “brazenly” and “illicitly” attempting to pilfer its AI capabilities. Anthropic alleged that Alibaba had orchestrated what it termed the “largest known distillation attack” on the company to date. For context, AI distillation involves a larger, more complex model teaching a smaller model to mimic its behavior, a process that can be sensitive if unauthorized.
Anthropic’s terms of service explicitly prohibit companies from “adversarial nations,” including China, from utilizing its models. In compliance with this directive, Alibaba employees are now mandated to uninstall all Anthropic models and agent products. They are instead required to utilize the Chinese company’s proprietary AI assistant, Qoder. Both Alibaba and Anthropic have declined to comment on the matter.
Alibaba’s decisive action coincides with a growing wave of online scrutiny directed at Anthropic within China. Posts on platforms like Reddit and GitHub have highlighted concerns about what appeared to be hidden code within Claude designed to detect user location, potentially indicating attempts to circumvent regional restrictions.
This situation has prompted Anthropic to reportedly take steps to close loopholes that have allowed Chinese firms to access Claude through third-country proxies. The Financial Times reported that Chinese fintech group Ant had provided employees with corporate Claude accounts accessed through the company’s intranet, routed via its Singapore-based entity. Similarly, while TikTok’s parent company ByteDance stated it does not directly facilitate Claude access, it implemented a reimbursement program allowing engineers to expense personal subscriptions, which could then be accessed via virtual private networks. Ant and ByteDance have not commented on these reports.
ByteDance’s reimbursement policy, initiated in early April, was framed as an initiative to encourage employees to “experience and learn” from a broader spectrum of AI products to enhance their skill sets, as per a person familiar with the matter. This strategy underscores a broader trend among major tech players to foster AI literacy and exploration among their workforce, even as regulatory and security concerns loom large.
The underlying technological and geopolitical tensions are significant. For U.S.-based AI developers like Anthropic, safeguarding intellectual property and ensuring compliance with national security directives are paramount. The alleged “distillation attacks” represent a direct threat to their competitive advantage and the immense R&D investments made in developing sophisticated AI models. On the other hand, Chinese tech giants are keen to leverage cutting-edge AI to maintain their global competitiveness, leading to a complex dance between innovation, access, and security. Alibaba’s move is a clear signal of its commitment to leveraging indigenous AI solutions while navigating these intricate global dynamics. The race for AI dominance is not just about algorithmic superiority but also about data security, intellectual property protection, and navigating an increasingly fragmented global tech landscape.
Original article, Author: Tobias. If you wish to reprint this article, please indicate the source:https://aicnbc.com/23458.html
