Representative Ro Khanna is advocating for a ban on recursive self-improving AI models, which can enhance their capabilities without human intervention. He supports the Ban Artificial Superintelligence Act, introduced on September 23, 2026, by Senator Bernie Sanders and Representative Greg Casar.
The proposed legislation includes severe penalties for violations, such as up to 20 years in prison and a corporate death penalty, which implies the forced dissolution of a company. The bill also calls for a pause on the development of superintelligent systems until a new Department of Artificial Intelligence is established.
Khanna proposes creating a federal AI Safety Agency, modeled after the FDA, to regulate frontier AI development. He also advocates for international agreements with China, suggesting verifiable inspections and accountability measures for major tech firms like Alibaba and DeepSeek, as well as US labs like Anthropic.
As of late September 2026, Khanna has not introduced his own standalone AI safety legislation. His current efforts are focused on supporting existing measures and promoting stricter commitments from laboratories and technology companies regarding the development of advanced systems.
