OpenAI has accused China’s DeepSeek of bypassing safeguards to distill and replicate U.S. AI models, raising concerns over technological competition.
OpenAI has told U.S. lawmakers that Chinese startup DeepSeek attempted to replicate its models through a technique known as distillation, according to a memo seen by Reuters. The company alleged “ongoing efforts to free-ride on the capabilities developed by OpenAI and other U.S. frontier labs.”
In the memo to a House committee, OpenAI said it observed accounts linked to DeepSeek employees circumventing access restrictions and using third-party routers to mask activity. DeepSeek did not immediately respond. OpenAI added it removes users suspected of attempting to distill its models.
NEWS NOW:
- Mayweather, Pacquiao rematch set for september
- Mexican cartel kingpin ‘El Mencho’ was taken out after authorities tracked mistress to secret love nest
- Anthropic reports AI model “distillation attacks” by rivals, DeepSeek and others
- Peter Thiel and other tech billionaires are publicly shielding their children from the products that made them rich