Major Chinese internet platforms have begun shutting down their consumer applications based on artificial intelligence chatbots, signaling a strategic shift: moving from independent chatbots to integrating AI features into existing products.
Major Chinese internet platforms have begun shutting down their consumer applications based on artificial intelligence chatbots, signaling a strategic shift: moving from independent chatbots to integrating AI features into existing products.
In recent months, ByteDance, Tencent, and Alibaba have discontinued or scaled back consumer chatbots, abandoning a market segment that attracted hundreds of millions of users just two years ago. These moves are part of a broader industry correction.
Each company's approach differs. Some are redirecting AI capabilities into core platform products, such as content feeds and messengers, while others are pausing operations to reassess product-market fit before relaunching with updated positioning.
Experts note that the initial hype surrounding chatbots has given way to a more deliberate integration phase. Companies that invested heavily in standalone chatbots are now questioning whether embedding AI capabilities directly into existing platforms with hundreds of millions of daily active users is more beneficial.
The competitive environment is also transforming. Open-source model providers and specialized AI startups are offering chatbot infrastructure at commodity prices, reducing the strategic value of owning a chatbot as a standalone consumer product. Simultaneously, the adoption of AI in the corporate sector is accelerating, directing investment funds toward B2B AI solutions rather than experimental consumer projects.
Likely to survive are those platforms that can find the clearest synergy between their AI functions and their core business models—be it social networks, content recommendation systems, e-commerce, or workplace productivity tools.
Alibaba will implement a ban for its employees from using Anthropic's artificial intelligence (AI) tools in work activities, effective this Friday, the 10th.
The Chinese company has placed Claude Code on an internal list of high-risk software. The justification provided is that the US-based company poses security threats due to the possibility of backdoors.
Consequently, Alibaba employees have been instructed to uninstall all Anthropic agents' models and products, and must migrate to the AI assistant developed internally by the Chinese company, called Qoder.
This determination comes after Anthropic sent correspondence in June to the U.S. Senate Committee on Banking, Housing, and Urban Affairs. In this document, the American company accused Alibaba of attempting to obtain its AI capabilities in a 'blatant' and 'illicit' manner, qualifying the move as 'the largest known distillation attack' against it to date.
Additionally, Anthropic's terms of service establish restrictions that prevent Chinese companies and other 'adversary nations' from using its models.
Alibaba's decision occurs alongside a series of negative reactions in China directed at Anthropic. These reactions were triggered by posts on Reddit and GitHub, where hidden codes designed to identify if users were located in a specific country were described.
On another front, the Financial Times reported on Friday, the 3rd, that Anthropic is taking steps to fix vulnerabilities that allowed Chinese companies to bypass restrictions and access Claude through third countries.
The British newspaper cited sources indicating that the fintech group Ant had provided employees with corporate Claude accounts, accessible via the company's intranet, which is linked to its headquarters in Singapore.
Regarding ByteDance, the owner of TikTok, although it does not facilitate access to Claude, it implemented a reimbursement system that allows engineers to register personal subscriptions as expenses using virtual private networks.
An individual familiar with the matter informed CNBC that the policy, disclosed on April 2nd, aims to encourage employees to 'experiment and learn' about a wider variety of AI products to enhance their skills. Neither Ant nor ByteDance issued statements regarding the Financial Times report, nor did Alibaba or Anthropic.
