Anthropic Accuses 3 Chinese Companies of Harvesting Its Data

NYT Homepage

by Cade Metz

February 23, 2026

AI-Generated Deep Dive Summary

Anthropic, a leading San Francisco-based artificial intelligence startup, has accused three prominent Chinese companies—DeepSeek, Moonshot, and MiniMax—of improperly harvesting data from its AI systems to accelerate their own chatbot development. The allegations center around the use of approximately 24,000 fraudulent accounts to generate over 16 million interactions with Anthropic’s Claude chatbot, potentially leveraging this data for distillation purposes. Distillation, a common practice in AI where one model learns from another, is typically restricted by terms of service agreements like those of Anthropic and OpenAI, which explicitly prohibit unauthorized data harvesting. The companies allegedly used these fraudulent accounts to engage with Claude, creating vast datasets that could be used to train their own chatbots. Anthropic highlights this as a significant breach of its terms of service, which also forbid the use of its technologies in China. The practice raises concerns about unfair competition and national security, particularly given the potential for such methods to advance AI capabilities in ways that bypass established guardrails designed to prevent harmful applications, such as bioweapons or mass surveillance tools. This issue is not isolated to Anthropic; OpenAI has also faced similar allegations from Chinese companies using its technology. In a memo addressed to the House Select Committee on China, OpenAI detailed how DeepSeek and others are employing obfuscated distillation techniques to "free-ride" on advancements made by U.S. companies. Both Anthropic and OpenAI view these practices as posing significant risks to global AI development and security. The controversy underscores broader tensions in the AI industry over data ownership, intellectual property, and ethical use. As Chinese companies continue to innovate rapidly in AI, questions about compliance with international norms and agreements will likely remain a focal point for tech firms and policymakers alike. This case highlights the delicate balance between fostering innovation and ensuring fair competition in the global AI landscape.

Verticals

newsgeneral

Originally published on NYT Homepage on 2/23/2026