Rep. Ro Khanna of California proposed Sunday that Congress establish a dedicated federal agency to regulate artificial intelligence development, modeled after the Food and Drug Administration's oversight structure for pharmaceuticals and medical devices.
Speaking on CBS News' "Face the Nation," Khanna argued the government should treat AI governance as a public health matter comparable to drug safety. The Democrat did not specify what regulatory powers such an agency would hold or what timeline he envisioned for its creation, but the proposal reflects growing concern among lawmakers that tech companies are advancing AI systems without adequate federal guardrails.
Khanna represents Silicon Valley and sits on the House Oversight Committee, positioning him as a bridge between tech industry interests and congressional regulation efforts. His call for an AI regulator places him among Democrats pushing for proactive governance as AI systems like ChatGPT gain rapid adoption. The proposal comes as regulators across federal agencies, including the Federal Trade Commission and Department of Commerce, have already begun examining AI's risks in areas ranging from job displacement to algorithmic bias to national security threats.
The FDA model Khanna referenced requires companies to submit products for approval before marketing them to consumers. Applying similar requirements to AI could mean developers must demonstrate safety and efficacy before releasing new systems. Such a framework would represent a dramatic shift from the current approach, where tech companies self-regulate through internal ethics boards while facing sporadic FTC enforcement actions for deceptive practices.
Creating an AI-specific agency would require legislation and face significant obstacles. Tech industry groups and libertarian-minded Republicans oppose new regulatory schemes, arguing they stifle innovation and give regulators too much power over cutting-edge technology. The industry has historically resisted prescriptive rules, preferring principles-based guidelines they can interpret flexibly.
Other lawmakers have proposed different regulatory approaches. Some favor giving existing agencies like the FTC stronger AI authority. Senator Chuck Schumer of New York introduced the SAFE Innovation Framework, which aims to balance oversight with innovation through a different model than traditional pre-market approval. The Biden administration released an AI executive order in October 2023 that directed agencies to develop standards and safety guidelines without creating a new agency.
Khanna's proposal also reflects anxieties about AI's development speed. OpenAI's rapid release of ChatGPT and subsequent improvements created pressure for regulation without giving policymakers time to understand the technology deeply. Safety researchers warn that large language models could pose risks if deployed without adequate safeguards, though those risks remain contested among experts.
The proposal faces questions about scope and jurisdiction. AI applications span healthcare, finance, employment, criminal justice, and defense. Each sector has different risks and existing regulatory bodies. An FDA-style agency might duplicate existing authority or create bureaucratic confusion.
Khanna's call signals that AI regulation will remain a Democratic priority heading into 2024 and beyond, even as business-friendly Republicans resist new regulatory structures. The debate essentially asks whether AI presents risks warranting pre-market approval or whether market forces and targeted enforcement suffice.