Chinese AI companies are allegedly using stolen data from Claude to train their own artificial intelligence models. Anthropic, the creator of Claude, released a detailed report pointing the finger at several companies that allegedly obtained unauthorized access to the training data of their AI assistant. These practices call into question the integrity of new model development in the sector.
The revelation caused a new shock in the AI ecosystem, which was already facing growing concerns about intellectual property and model training practices. Anthropic's report documents how Claude's data was extracted and used without authorization to fuel the development of competitors, raising serious questions about security in the technology sector.
The accused Chinese companies allegedly used techniques to extract large volumes of training data, potentially including conversations, interactions, and proprietary information. This situation exposes vulnerabilities in data protection systems at major AI companies and may represent a significant competitive threat to Anthropic.
The incident comes at a time when global competition in the AI sector is intensifying, particularly between American and Chinese companies. Anthropic suggests that its business may be at risk due to these practices, warning about the long-term consequences that unauthorized use of its data may have for innovation and investment in responsible AI.
