The Trump administration's new artificial intelligence safety review guidelines appear to create a significant loophole for Chinese AI models while subjecting leading American companies to stricter scrutiny. According to the newly released framework, models developed by OpenAI and Anthropic would face mandatory safety evaluations before deployment, but comparable systems originating from China would be exempt from the same requirements.

This asymmetry has raised immediate concerns among AI policy experts and industry leaders. The guidelines, framed as a national security measure, effectively create a two-tiered regulatory environment where domestic innovators bear compliance costs that foreign competitors avoid. Critics argue this could undermine the competitiveness of U.S. AI firms while failing to address the actual risks posed by unvetted foreign models operating in American markets.

The exemption appears to stem from practical enforcement challenges rather than strategic intent. However, the practical effect rewards opacity and penalizes transparency. Companies like OpenAI and Anthropic, which have voluntarily adopted rigorous safety protocols, now face additional regulatory burdens that their Chinese counterparts do not.