Anthropic AI has accused Chinese technology giant Alibaba of a large-scale effort to extract capabilities from its Claude AI models.
The allegation has escalated tensions in the global AI race and raised fresh concerns about intellectual property, model security and cross-border technology competition.
According to a letter sent by Anthropic to US lawmakers, operators allegedly linked to Alibaba and its AI research arm Qwen used nearly 25,000 fraudulent accounts to interact with Claude between 22 April and 5 June 2026.
The company claims the activity resulted in more than 28.8 million exchanges with the chatbot, making it the largest known incident of its kind targeting Anthropic to date.
Also Read: Anthropic AI Pulls Advanced Models Worldwide After US Restricts Foreign Access
What Anthropic Says Happened
Anthropic alleges that the campaign was designed to conduct what the AI industry calls “distillation”, a process in which a smaller or less capable AI system learns from the responses of a more advanced model.
The company claims the activity focused on some of Claude’s most valuable capabilities, including software development assistance, advanced reasoning and the ability to handle complex multi-step tasks. Anthropic argues that such efforts allow competitors to accelerate AI development without bearing the full cost of research, training and computing infrastructure.
The letter reportedly described the operation as an attempt to replicate capabilities similar to those of Anthropic’s latest AI systems. It warned that such practices could narrow the technology gap between leading US and Chinese AI developers.
Also Read: Independent AI: Anthropic Says Claude AI Can Now Improve Its Own Code
Growing Scrutiny Around AI Model Security
The allegations come as governments and AI companies worldwide face increasing pressure to safeguard advanced AI systems from misuse, unauthorised access and model replication.
Anthropic has previously raised similar concerns involving other Chinese AI firms, including DeepSeek, Moonshot AI and MiniMax, which it accused earlier this year of using large-scale account networks to access Claude outputs.
The dispute highlights how the battle for AI leadership is increasingly shifting beyond model performance to issues of access, security and intellectual property. As governments tighten oversight of advanced AI technologies, scrutiny of how companies develop, train and protect their models is likely to intensify in both the US and China.
Samarjit Kaur is a journalist and communications professional covering technology & emerging digital trends. With a focus on clarity and context, she reports on developments shaping industries and governance. When not reporting, she chooses to plug-in and relax on her playlists and plan her next bucket-list trips!
