Skip to content
Search

Latest Stories

Top Stories

AI Is Here. Our Laws Are Stuck in the Past.

Opinion

Closeup of Software engineering team engaged in problem-solving and code analysis

Closeup of Software engineering team engaged in problem-solving and code analysis.

Getty Images, MTStock Studio

Artificial intelligence (AI) promises a future once confined to science fiction: personalized medicine accounting for your specific condition, accelerated scientific discovery addressing the most difficult challenges, and reimagined public education designed around AI tutors suited to each student's learning style. We see glimpses of this potential on a daily basis. Yet, as AI capabilities surge forward at exponential speed, the laws and regulations meant to guide them remain anchored in the twentieth century (if not the nineteenth or eighteenth!). This isn't just inefficient; it's dangerously reckless.

For too long, our approach to governing new technologies, including AI, has been one of cautious incrementalism—trying to fit revolutionary tools into outdated frameworks. We debate how century-old privacy torts apply to vast AI training datasets, how liability rules designed for factory machines might cover autonomous systems, or how copyright law conceived for human authors handles AI-generated creations. We tinker around the edges, applying digital patches to analog laws.


This constant patching creates what we might call "legal tech debt." Imagine trying to run sophisticated AI software on a computer from the 1980s—it might technically boot up, but it will be slow, prone to crashing, and incapable of performing its intended function. Similarly, forcing AI into legal structures designed for a different technological era means we stifle its potential benefits while failing to adequately manage its risks. Outdated privacy rules hinder the development of AI for public good projects; ambiguous liability standards chill innovation in critical sectors; fragmented regulations create uncertainty and inefficiency.

Allowing this legal tech debt to accumulate isn't just about missed opportunities; It breeds public distrust when laws seem irrelevant to lived reality. It invites policy chaos, as seen with the frantic, often ineffective, attempts to regulate social media after years of neglect. It risks a future where transformative technology evolves haphazardly, governed by stopgap measures and reactive panic rather than thoughtful design. With AI, the stakes are simply too high for such recklessness.

We need a fundamentally different approach. Instead of incremental tinkering, we need bold, systemic change. We need to be willing to leapfrog—to bypass outdated frameworks and design legal and regulatory systems specifically for the age of AI.

What does this leapfrog approach look like? It requires three key shifts in thinking:

First, we must look ahead. Policymakers and experts need to engage seriously with plausible future scenarios for AI development, learning from the forecasting methods used by technologists. This isn’t about predicting the future with certainty but about understanding the range of possibilities—from accelerating breakthroughs to unexpected plateaus—and anticipating the legal pressures and opportunities each might create. We need to proactively identify which parts of our legal infrastructure are most likely to buckle under the strain of advanced AI.

Second, we must embrace fundamental redesign. Armed with foresight, we must be willing to propose and implement wholesale reforms, not just minor rule changes. If AI requires vast datasets for public benefit, perhaps we need entirely new data governance structures—like secure, publicly accountable data trusts or commons—rather than just carving out exceptions to FERPA or HIPAA. If AI can personalize education, perhaps we need to rethink rigid grade-based structures and accreditation standards, not just approve AI tutors within the old system. This requires political courage and a willingness to question long-held assumptions about how legal systems should operate.

Third, we must build in adaptability. Given the inherent uncertainty of AI’s trajectory, any new legal framework must be dynamic, not static. We need laws designed to evolve. This means incorporating mechanisms like mandatory periodic reviews tied to real-world outcomes, sunset clauses that force reconsideration of rules, specialized bodies empowered to update technical standards quickly, and even using AI itself to help monitor the effectiveness and impacts of regulations in real-time. We need systems that learn and adapt, preventing the accumulation of new tech debt.

Making this shift won't be easy. It demands a new level of ambition from our policymakers, a greater willingness among legal experts to think beyond established doctrines, and broader public engagement on the fundamental choices AI presents. But the alternative—continuing to muddle through with incremental fixes—is far riskier. It’s a path toward unrealized potential, unmanaged risks, and a future where technology outpaces our ability to govern it wisely.

AI offers incredible possibilities but realizing them requires more than just brilliant code. It requires an equally ambitious upgrade to our legal and regulatory operating system. It’s time to stop patching the past and start designing the future. It’s time to leapfrog.

Kevin Frazier is an AI Innovation and Law Fellow at Texas Law and Author of the Appleseed AI substack.

Read More

Fear of AI Makes for Bad Policy
Getty Images

Fear of AI Makes for Bad Policy

Fear is the worst possible response to AI. Actions taken out of fear are rarely a good thing, especially when it comes to emerging technology. Empirically-driven scrutiny, on the other hand, is a savvy and necessary reaction to technologies like AI that introduce great benefits and harms. The difference is allowing emotions to drive policy rather than ongoing and rigorous evaluation.

A few reminders of tech policy gone wrong, due, at least in part, to fear, helps make this point clear. Fear is what has led the US to become a laggard in nuclear energy, while many of our allies and adversaries enjoy cheaper, more reliable energy. Fear is what explains opposition to autonomous vehicles in some communities, while human drivers are responsible for 120 deaths per day, as of 2022. Fear is what sustains delays in making drones more broadly available, even though many other countries are tackling issues like rural access to key medicine via drones.

Keep ReadingShow less
A child looking at a smartphone.

With autism rates doubling every decade, scientists are reexamining environmental and behavioral factors. Could the explosion of social media use since the 1990s be influencing neurodevelopment? A closer look at the data, the risks, and what research must uncover next.

Getty Images, Arindam Ghosh

The Increase in Autism and Social Media – Coincidence or Causal?

Autism has been in the headlines recently because of controversy over Robert F. Kennedy, Jr's statements. But forgetting about Kennedy, autism is headline-worthy because of the huge increase in its incidence over the past two decades and its potential impact on not just the individual children but the health and strength of our country.

In the 1990s, a new definition of autism—ASD (Autism Spectrum Disorder)—was universally adopted. Initially, the prevalence rate was pretty stable. In the year 2,000, with this broader definition and better diagnosis, the CDC estimated that one in 150 eight-year-olds in the U.S. had an autism spectrum disorder. (The reports always study eight-year-olds, so this data was for children born in 1992.)

Keep ReadingShow less
Tech, Tribalism, and the Erosion of Human Connection
Ai technology, Artificial Intelligence. man using technology smart robot AI, artificial intelligence by enter command prompt for generates something, Futuristic technology transformation.
Getty Images - stock photo

Tech, Tribalism, and the Erosion of Human Connection

One of the great gifts of the Enlightenment age was the centrality of reason and empiricism as instruments to unleash the astonishing potential of human capacity. Great Enlightenment thinkers recognized that human beings have the capacity to observe the universe and rely on logical thinking to solve problems.

Moreover, these were not just lofty ideals; Benjamin Franklin and Denis Diderot demonstrated that building our collective constitution of knowledge could greatly enhance human prosperity not only for the aristocratic class but for all participants in the social contract. Franklin’s “Poor Richard’s Almanac” and Diderot and d’Alembert’s “Encyclopédie” served as the Enlightenment’s machines de guerre, effectively providing broad access to practical knowledge, empowering individuals to build their own unique brand of prosperity.

Keep ReadingShow less
The limits of free speech protections in American broadcasting

FCC Chairman Brendan Carr testifies in Washington on May 21, 2025.

The limits of free speech protections in American broadcasting

The chairman of the Federal Communications Commission is displeased with a broadcast network. He makes his displeasure clear in public speeches, interviews and congressional testimony.

The network, afraid of the regulatory agency’s power to license their owned-and-operated stations, responds quickly. They change the content of their broadcasts. Network executives understand the FCC’s criticism is supported by the White House, and the chairman implicitly represents the president.

Keep ReadingShow less