Did DeepSeek steal OpenAI data for training? Microsoft begins probe: Report
Microsoft has launched an investigation into a group linked to Chinese AI startup DeepSeek for allegedly unauthorized data extraction from OpenAI’s API. This breach raises critical questions about the security vulnerabilities inherent in open APIs and the potential for industrial espionage within the rapidly evolving artificial intelligence sector. The incident underscores significant challenges in protecting proprietary algorithms and large language model weights from unauthorized access or "distillation." It highlights the tension between open innovation and the need for robust security measures to prevent the illicit copying of advanced technological assets by competitors. This situation is highly relevant to open_data advocates, as it demonstrates the real-world risks associated with sharing powerful AI models. It suggests that while openness drives progress, it must be balanced with strict access controls to safeguard intellectual property and maintain trust in the global AI infrastructure.
Source: hindustantimes.comPublished on 2025-01-30