BlockBeats News, June 26th, Serenity tweeted about the phenomenon of US companies extensively using Chinese AI models. Citing a UBS report, Serenity pointed out that many enterprises have started routing simple tasks to cheaper Chinese open-source models, with some teams spending up to $35,000 in tokens per month, exceeding 200%. As a result, the revenue of high-end models is under pressure. Serenity confirmed the accuracy of the UBS report and attributed it to a dilemma—where the market naturally opts for the cheapest solution, with Chinese models like DeepSeek being much cheaper than Gemini, OpenAI, and Anthropic.

Serenity also stated that the Trump administration's earlier suspension of Fable/Mythos access was a correct move, as the top-tier models are at high risk of repeated distillation, and models approaching AGI levels should have their access thresholds raised. Serenity believes the US needs to do two things: further develop models adept at cost-effective reasoning and impose bank-level identity authentication for AI model access.