Home / Technology / OpenAI Warns of Chinese AI 'Free-Riding'
OpenAI Warns of Chinese AI 'Free-Riding'
13 Feb
Summary
- OpenAI alerted U.S. lawmakers about DeepSeek's alleged model replication.
- DeepSeek employees reportedly used code to access U.S. AI models.
- China's AI models are accused of cutting corners on safety.
OpenAI has issued a warning to U.S. lawmakers concerning the activities of Chinese AI startup DeepSeek. The company alleges that DeepSeek has been attempting to "free-ride" on the advancements made by OpenAI and other major U.S. artificial intelligence laboratories. These alleged efforts involve replicating and utilizing established AI models for DeepSeek's own training purposes.
According to a memo shared with the U.S. House Select Committee on Strategic Competition between the U.S. and the Chinese Communist Party, OpenAI detailed methods allegedly employed by DeepSeek employees. These include circumventing access restrictions and using code to programmatically access and distill outputs from U.S. AI models. OpenAI stated it proactively removes users attempting such actions.
Hangzhou-based DeepSeek gained market attention last year with AI models that rivaled U.S. offerings. However, OpenAI expressed concerns that Chinese large language models are "actively cutting corners when it comes to safely training and deploying new models."
