The Trump administration is considering requiring US companies to publicly disclose when they use Chinese artificial-intelligence models, according to Reuters. The proposal would not immediately ban those systems. Instead, companies could be asked to publish the information on their websites, making model origin part of the public record. Reuters reported that Treasury officials and the White House Office of Science and Technology Policy discussed the idea. The move reflects a widening debate over whether AI procurement should be treated as a technology choice, a national-security question, or both.

US and Chinese flags at a technology policy event

The report cited Airbnb as an example because the company uses Alibaba’s Qwen model for an AI customer-service agent. Airbnb said most of its AI activity runs on US-origin models and that its limited China-origin use is handled through approved US-based providers. That distinction shows why disclosure rules could become technically complicated. Companies may rely on several models, host open-weight systems themselves, or access them through cloud services. A useful rule would need to define what counts as use and which parts of a technology stack must be disclosed.

AI technology showcase with robotics and smart appliances

Chinese open-weight models are attracting attention because they can be cheaper than proprietary systems and perform well enough for selected business tasks. Reuters said models such as Kimi, DeepSeek, and Qwen have gained traction, while US lawmakers have asked companies about their use. Open-weight access can support experimentation and lower costs, but it can also create questions about data handling, software updates, model provenance, and dependence on a foreign ecosystem. Those concerns are pushing organizations to document more than the name of the application they deploy.

AI Zone floor at a digital technology exhibition

For businesses in the Philippines, the developing US debate is a reminder to make AI governance practical and model-specific. Procurement teams should know where a model comes from, where prompts and outputs are processed, which provider controls updates, and what happens if a service becomes restricted. Public disclosure may remain only a proposal, but the underlying issue is already relevant to banks, governments, software companies, and customer-service operators. Clear inventories and vendor reviews can reduce operational surprises without treating every foreign model as automatically unsafe.