December 2, 2024
The landscape of artificial intelligence (AI) transformed dramatically on November 30, 2022, when OpenAI released ChatGPT, a large language model that captured global attention with its ability to engage in natural conversations, write creative content, generate computer code, and solve complex problems. While OpenAI was already well-known in tech circles for innovations like GPT-3 and DALL-E, ChatGPT’s launch marked a watershed moment that sparked widespread public interest in AI. In response, technology giants, including Google, Microsoft, Meta, and Anthropic, accelerated their AI development, releasing sophisticated models like Gemini, Claude, and Llama. This AI revolution has reshaped numerous sectors, from education and healthcare to software development and creative industries, fundamentally changing how we work, learn, and interact with technology.
The rapid expansion of AI technology has prompted policymakers, technologists, and advocacy groups to scrutinize the technology’s societal implications, from workforce transformation to algorithmic fairness and mental health impacts. In response, legislators have introduced hundreds of bills addressing challenges, from ensuring election integrity by banning the use of AI in election advertisements to criminalizing the use of AI to create child pornography. Additionally, several states like Colorado and Connecticut have imposed comprehensive AI regulations that would establish inhospitable governance frameworks for AI development and deployment.
As we approach 2025, the surge of AI-related legislation shows no signs of slowing, putting state lawmakers at a critical juncture. State lawmakers entering the 2025 legislative session must carefully consider how to craft regulations that protect citizens while fostering innovation and maintaining American technological leadership.
While state legislatures have taken the lead on regulating AI, this decentralized approach risks creating a patchwork of inconsistent standards across the country. These varying and potentially conflicting regulations not only create uncertainty for consumers about their rights but also increase compliance costs for businesses operating across multiple jurisdictions. Given that AI companies typically operate nationwide, Congress would be better positioned to establish a comprehensive governance framework that creates clear and consistent rules across all states, rather than leaving regulation to individual states.
Beyond the jurisdictional challenges, lawmakers must also grapple with complex technical issues, particularly the question of algorithmic fairness. While historical AI training data has often underrepresented certain communities, creating new AI-specific anti-discrimination laws may be redundant given that existing civil rights statutes already prohibit discriminatory outcomes regardless of the technology involved. Additionally, the market already provides a critical incentive for developers to use more representative training data that better reflects America’s demographic makeup. As more representative training data is used, issues of algorithmic discrimination should become less pronounced.
While addressing discriminatory outcomes remains crucial, an equally fundamental challenge lies in the very language used to regulate AI systems. Definitions have become a critical component of AI legislation. As lawmakers have sought to legislate AI, they have often included overly broad definitions that risk capturing other technologies beyond the intended regulation. For example, many states have employed definitions that target algorithmic outputs that could regulate Netflix recommendations. Employing such an overly broad definition is problematic because it creates confusion as to what is and is not subjected to regulation and may be more encompassing than needed. Instead, lawmakers should ensure any definition is narrowly tailored.
As AI technology continues its rapid evolution, policymakers in state capitols across the country will face the complex challenge of developing appropriate regulatory frameworks. While state-level initiatives demonstrate a commitment to addressing AI’s societal impacts, a federal approach may better serve both consumers and businesses by establishing consistent nationwide standards. Moving forward, successful AI regulation will require carefully crafted definitions, recognition of existing legal frameworks, and a balanced approach that protects citizens while fostering innovation. Most importantly, policymakers must remain adaptable, as today’s regulatory solutions may need to evolve alongside tomorrow’s technological advances.
