ChatGPT maker OpenAI lately accused Chinese AI startup DeepSeek of using its proprietary models for the training of its AI system. It claims having evidence that the competitive AI chatbot has been in a practice called \u201cdistillation,\u201d which means the developers used outputs from larger AI models to enhance smaller ones.<\/p>\n
The issue first came to light when the security team of Microsoft noticed unusual activity last year. They detected that individuals linked to DeepSeek were extracting large amounts of data from the API of OpenAI. The distillation is a common AI technique and OpenAI argues that DeepSeek violated its terms of service. OpenAI strictly prohibits using its models to develop competing systems. It has now blocked the suspected DeepSeek accounts from accessing its API.<\/p>\n
DeepSeek recently gained attention for developing its AI model named R1 at a much lower cost than industry norms. The company claims that it trained the model using 2,048 Nvidia H800 GPUs for just $5.6 million. The model reportedly has 671 billion parameters and this makes it one of the more advanced AI models developed in China.<\/p>\n
The controversy has now also been reached the US government. David Sacks, an AI advisor at the White House, said that there is substantial evidence that DeepSeek has extracted knowledge from the models of OpenAI.<\/p>\n
OpenAI has also raised concerns that Chinese firms are attempting to replicate the US AI technologies.<\/p>\n
However, the incident now highlights the growing tension over AI development, intellectual property and global competition in the industry.<\/p>\n","protected":false},"excerpt":{"rendered":"