The Donald Trump administration is preparing a strategy to limit access to advanced Chinese artificial intelligence (AI) models. According to information from Axios, these steps are motivated by cybersecurity concerns and aim to strengthen the positions of leading American technology companies.
The rise of the Chinese open-source model "Kimi" has changed market dynamics. American firms are increasingly turning to Chinese solutions due to their lower prices and high capabilities, which are now nearly catching up to American alternatives. This, however, has triggered a reaction in the White House.
Planned measures include:
- Licensing restrictions: The Department of Commerce is considering including Chinese AI laboratories on a list of restricted entities.
- Executive orders: It is possible that liability will be introduced for American companies that host Chinese models, requiring them to guarantee full security against potential vulnerabilities.
- Open-source regulations: Rules are being discussed that would make it more difficult to use foreign open-source models in the US.
Instead of a total ban, the government may opt for a "softer" strategy, involving strict rules for public procurement and restrictions based on national security risks. The goal is not only to limit China's influence but also to stimulate the creation of a competitive American open-source ecosystem.
Critics of these measures, including AI advisor David Sacks, warn that excessive regulation could eliminate competition and only benefit the largest laboratories like OpenAI and Anthropic, which support stricter licensing frameworks.