Focal Foto
This report was shared by Bloomberg, referencing a letter that OpenAI sent to the United States House Select Committee on Strategic Competition between the United States and the Chinese Communist Party.
OpenAI stated that DeepSeek employed so-called "distillation" methods, where one artificial intelligence model is trained using the outputs of another to gain comparable abilities. In a document provided to Congress, OpenAI mentioned that the majority of hostile distillation activities observed on their platform seem to come from China, with some instances also occurring in Russia.
ChatGPT's creator claims that DeepSeek is continuously trying to take advantage of the abilities created by OpenAI and other top US research institutions. The company also mentioned that it found "new, secret ways" to bypass the security measures put in place by OpenAI to stop the improper use of its model results. OpenAI thinks that DeepSeek's approach is getting more advanced.
source: bloomberg.com
OpenAI stated that DeepSeek employed so-called "distillation" methods, where one artificial intelligence model is trained using the outputs of another to gain comparable abilities. In a document provided to Congress, OpenAI mentioned that the majority of hostile distillation activities observed on their platform seem to come from China, with some instances also occurring in Russia.
ChatGPT's creator claims that DeepSeek is continuously trying to take advantage of the abilities created by OpenAI and other top US research institutions. The company also mentioned that it found "new, secret ways" to bypass the security measures put in place by OpenAI to stop the improper use of its model results. OpenAI thinks that DeepSeek's approach is getting more advanced.
source: bloomberg.com