Restrain Chinese companies? Uncommon cooperation among AI giants in the United States.

The three American artificial intelligence giants, originally competitors, are now seemingly cooperating in a rare move to curb Chinese companies, reflecting the increasingly intense technological competition between China and the United States in the field of AI. Bloomberg reported on the 6th, citing insiders, that OpenAI, Anthropic, and Google under Alphabet have begun collaborating to share information through the "Frontier Model Forum," trying to curb Chinese competitors from extracting results from advanced AI models in the United States to gain an advantage in global AI competition. The so-called "Frontier Model Forum" is an industry non-profit organization jointly established by the three companies and Microsoft in 2023, aimed at identifying so-called "adversarial distillation" behaviors that violate terms of service. Bloomberg reported that these companies are concerned that some users, especially Chinese users, are developing counterfeit versions of their products, which could compete for customers at lower prices while posing security risks. OpenAI confirmed participation in the information sharing conducted through the "Frontier Model Forum" and mentioned a recent memorandum submitted to the US Congress, accusing the Chinese company DeepSeek of "freely utilizing the capabilities developed by OpenAI and other advanced labs in the United States." Google, Anthropic, and the aforementioned forum all declined to comment. Bloomberg reported that the rare cooperation of the three American AI companies highlights the seriousness with which American AI companies are addressing these issues.