According to Decrypt, Chinese internet authorities have launched an official investigation into domestic artificial intelligence firms DeepSeek and Moonshot AI. This action follows serious accusations made by US-based company Anthropic regarding unauthorized data transmission practices involving its flagship Claude model.
The controversy centers on claims that both investigated entities secretly routed millions of user conversations directly to Anthropic’s systems without proper consent or notification. These intercepted exchanges were allegedly used for training DeepSeek and Moonshot proprietary models, raising significant concerns about intellectual property violations within the rapidly expanding global AI sector.
This incident has intensified debates surrounding ethical data usage standards across international borders as tech companies navigate complex regulatory landscapes. The probe represents a direct intervention by Chinese regulators who are scrutinizing alleged misconduct that potentially undermines fair competition principles in artificial intelligence development worldwide.
The investigation highlights growing tensions between domestic and foreign AI corporations over proprietary model training methods and the treatment of user data during cross-border collaborations. As Anthropic’s accusations gain traction, industry stakeholders must reconsider how they approach sensitive information sharing while developing advanced language models that serve diverse global markets effectively.
