OpenAI has stated that the Chinese artificial intelligence company DeepSeek trained its large language model (LLM) using OpenAI's proprietary data. This claim emerged following an analysis by OpenAI researchers who identified similarities between DeepSeek's LLM and OpenAI's GPT-2 model. The researchers found that DeepSeek's model exhibited characteristics suggesting it was trained on a dataset that included OpenAI's data, potentially violating OpenAI's terms of service. This revelation raises concerns about data privacy and intellectual property within the rapidly evolving AI landscape, particularly regarding the unauthorized use of training data by competing entities.