For decades, the image of government regulation has been synonymous with dusty, overflowing binders and legal jargon. State regulatory codes tend to accumulate year after year, sometimes contradicting themselves and complicating life for those navigating the process. This accumulation of “red tape” creates significant burdens for small business owners trying to remain compliant, citizens trying to understand their rights, and even regulators trying to enforce outdated mandates.
Traditional regulatory reform has been a difficult task: manual reviews by human staffers attempting to read, analyze, and cross-reference thousands of pages of text. It was slow, expensive, and prone to error.
Today, however, a new tool is entering the civic arena: Artificial Intelligence, or AI. Most of us have heard of AI by now–as the creator of social media videos and memes, as the magical creator of new Randy Travis songs, or as the generator of instant research. If you’ve got a smartphone or have used an internet search engine, you’ve used AI without really thinking about it.
Natural Language Processing (NLP) is a field of artificial intelligence that enables computers to understand, interpret, and generate human language, including text and speech. It uses a combination of statistical methods, machine learning, and linguistic rules to process unstructured data and enable human-computer interaction through applications such as machine translation, chatbots, and voice recognition. Natural Language Processing is transforming regulatory reform from a sluggish manual process into a dynamic, data-driven exercise. By deploying AI to scan vast repositories of legal text, states are modernizing their codes with unprecedented speed and accuracy.
How AI “Reads” the Law
The primary challenge with regulatory code is its sheer volume and complexity. A human lawyer might take days to analyze a single chapter for inconsistencies thoroughly. An AI model trained on legal text can scan an entire state code in hours.
These AI tools are not making policy decisions; rather, they are performing diagnostic work. They use advanced algorithms to:
-
Identify Outdated Language: AI can flag references to mostly-obsolete technologies (like fax machines or typewriters) or defunct agencies.
-
Spot Duplication and Contradiction: The software can identify where two different regulations govern the same activity, potentially offering contradictory instructions.
-
Measure Regulatory Burden: By counting restrictive terms—such as “shall,” “must,” “required,” or “prohibited”—AI gives policymakers a quantitative “heat map” of where the regulatory burden is heaviest. The Mercatus Center has already done several studies for the Platte Institute using this sort of data, including this one just last year.
-
Improve Readability: AI tools can analyze the linguistic complexity of a regulation, helping officials rewrite rules in plain English so they are accessible to non-lawyers.
Laboratories of Innovation: States Leading the Charge
While the federal government is exploring these tools, U.S. states—the traditional “laboratories of democracy”—are currently at the forefront of applying AI to regulatory reform. They are finding that technology is essential for keeping their code lean and economically competitive.
Idaho: The Pioneer of Simplification. Idaho has been a national leader in regulatory cleanup. Through a concerted, multi-year effort involving rigorous review processes aided by technology, Idaho surpassed South Dakota to become the least regulated state in the nation. They used data analytics tools to track word counts and restricted terms, setting a baseline that enabled them to aggressively cut thousands of pages of outdated code. Their success proved that data-driven transparency is the first step toward meaningful reform.
Virginia: Targeting Regulatory Burdens. More recently, Virginia launched innovative pilot programs designed to reduce the regulatory compliance burden on businesses. The state has utilized AI platforms to scan its administrative code, specifically looking for duplicative or overly burdensome requirements. By using AI to highlight the most complex sections of their code, Virginia regulators can prioritize their human review efforts on the areas that will yield the biggest economic benefit. The goal is not just fewer words, but smarter, more efficient governance.
Utah: A Sandbox for Innovation. Utah has embraced a forward-thinking approach to governance through its Office of Regulatory Relief. While known for its “regulatory sandboxes” (which allow businesses to test new concepts under eased regulations), Utah is also actively exploring how AI and machine learning can modernize its administrative code. By adopting a tech-first mindset, the state aims to create a regulatory environment that keeps pace with modern business speeds rather than hindering them with analog-era rules.
The Human Element Remains Crucial
While the potential of AI in regulatory reform is immense, it’s important to recognize its limitations. AI is a powerful flashlight in a dark warehouse; it illuminates the mess, but it doesn’t clean it up.
There are risks associated with deploying these tools. If the data fed into the AI is flawed, the analysis will be skewed. Furthermore, there is the “black box” problem—it is not always clear how an AI model reaches its conclusions.
Therefore, the most successful state models should keep humans firmly in the loop. For instance, the AI might identify a potential conflict between a 1995 environmental rule and a 2023 housing rule, but a qualified human policymaker must decide which rule takes precedence and how to resolve the dispute. The goal is augmented intelligence, where technology handles the immense cognitive load of data processing, freeing up public servants to focus on strategy and judgment.
Conclusion
The digitization of regulatory reform is not merely an administrative upgrade; it is a necessary evolution for modern governance. By harnessing the power of AI, states can move from a reactive posture—where rules accumulate unnoticed until they cause a crisis—to a proactive one. The result is a regulatory environment that is leaner, more transparent, and easier for citizens and businesses to navigate.
