US agencies say Moonshot's Kimi was trained on American models

Six Chinese AI firms, including Moonshot, maker of the Kimi model that rattled markets earlier this year, have been accused by U.S. agencies of systematically extracting capabilities from American AI models, according to CoinDesk. The allegation targets a pattern of reverse-engineering or training on U.S. systems rather than capturing capability through independent research and development.

No technical specifics, evidence details, or any formal agency action were disclosed in the initial report, and the firms have not publicly responded to the allegations.

• Moonshot is named as one of six Chinese AI firms in the accusation.
• The claim centers on U.S. models being used to source capabilities.
• Limited detail is available beyond the agencies' initial public statements.