Content area
The emergence of OpenAI's ChatGPT has marked the dawn of a new Artificial Intelligence (AI) era, unleashing a wave of user excitement, moral panic, and apocalyptic warnings. Yet, these concerns have done little to slow the breakneck pace of AI development and deployment.
Should the United States adopt the European Union's top-down, legislation-first approach, or follow the United Kingdom's bottom-up, court-driven model? This research draws from historical legislative parallels in the U.S. By examining past regulatory frameworks for disruptive technologies, including electricity, automobiles, the telephone, broadband internet, nuclear technology, cryptocurrency, commodity futures, modern pharmaceuticals, and social media, we identify key legislative and institutional lessons.
Three regulatory archetypes emerge as particularly instructive:
• Commodity futures and cryptocurrency regulation illustrate a principles-based, agency-coordinated, and SRO-supported model, offering adaptability and sectoral alignment.
• Modern pharmaceutical regulation offers a roadmap for FDA-style pre-market registration, clinical trials, developer-centered liability, and post-deployment surveillance.
• Nuclear governance underscores the necessity of global coordination for the peaceful deployment of AI technologies amid fierce rivalry among leading AI powers.
From these precedents, the dissertation develops a contemporary framework centered on key regulatory enablers: registration, monitoring, and enforcement. It introduces the AI Pentad, a model identifying five foundational components of AI systems: humans and organizations, algorithms, data, compute, and energy. Integrating this with the enablers yields the CHARME2D Model, a comprehensive structure to guide, construct, and assess AI regulatory regimes.
Finally, the dissertation expands its lens to the international arena, drawing on nuclear regulation as a case study in multilateral governance. International AI governance is currently stalled, due in large part to the intensifying rivalry between the United States and China. We propose a strategic pivot to establish an International AI Agency (IAIA), resembling the International Atomic Energy Agency (IAEA), but excluding the United States and China from the onset. The IAIA would fulfill a role the U.S. might have led under different geopolitical conditions—but is now constrained by its rivalry with China. The United States should nonetheless support the IAIA as a constructive and peaceful third force.