Kratsios AI Plan Ties Open Weights to Regulatory Exemption

Michael Kratsios has proposed a conditional US AI regulatory framework that exempts open-weight models from oversight while strictly regulating closed systems, according to comments analyzed by Matan Cohen. Cohen stated this approach favors domestic open model development but warned that US companies cannot legally distill existing models without facing copyright infringement lawsuits. He argued that removing this legal barrier is essential for American AI competitiveness against foreign rivals. Cohen further noted that absent legal reform, investors require clearer revenue pathways to fund open-weight ventures. The proposal effectively creates a two-tier system where transparency grants regulatory relief. However, intellectual property disputes regarding training data remain a significant obstacle to this strategy's success. Industry stakeholders now await clarification on how distillation liability will be addressed under potential new guidelines.