Law as Code: The AI Power Shift
In this episode of New Law Order, the limited series co-hosted by Joel Cohen and Yale Law Professor John Morley, Yale Law professor and legal philosopher Scott Shapiro reframes the AI conversation away from productivity and toward power. Artificial intelligence is often marketed as a tool that makes lawyers faster. Shapiro argues that the deeper shift is structural. If law is a system for coordinating behavior at scale, then machines that can interpret rules, test edge cases, and generate persuasive legal analysis may change who can navigate that system—and who gets overwhelmed by it. (Watch free on Vimeo.)
Law as a Social Technology
Shapiro begins at first principles. Law is not merely a body of rules; it is a social technology—an engineered system designed to guide human conduct. Legal institutions assume human judgment, human limitations, and human-scale reasoning. When machines begin operating inside that system, the assumptions embedded in it begin to strain.
The result is not simply efficiency. It is a reallocation of institutional advantage. Tools that interpret dense regulatory schemes or simulate litigation outcomes can empower individuals and small firms. They can also concentrate leverage in the hands of sophisticated actors who know how to deploy them strategically.
Compliance Engine vs. Exploitation Engine
One of the episode’s central tensions is what Shapiro describes as a darker symmetry: the same AI systems that help organizations comply with the law can also help them evade it.
AI can:
-
Stress-test regulations for weak points
-
Identify contractual ambiguities
-
Generate arguments at scale
-
Simulate enforcement thresholds
In other words, models can act as compliance engines—or exploitation engines. The distinction depends less on the tool and more on the incentives of the user. That dual-use reality raises uncomfortable questions about whether AI will rebalance justice or simply accelerate legal arbitrage.
Generative Models vs. Rule-Based Systems
Shapiro draws a sharp distinction between generative systems and rule-based reasoning.
Generative models can produce fluent, authoritative-sounding legal analysis while being intentionally or accidentally wrong. The risk is epistemic: hallucinations, confident errors, and manufactured authority.
Rule-based systems present a different problem. When logic is technically correct but too complex for humans to follow, legitimacy begins to erode. At that point, legal reasoning risks becoming an appeal to authority—not because it lacks logic, but because it exceeds human comprehension.
The conversation presses on a critical question: what does transparency mean when legal reasoning is no longer human-readable?
When Legal Reasoning Stops Being Human
Legal systems derive legitimacy not only from outcomes but from process—from the ability of citizens, lawyers, and judges to understand the reasoning behind decisions. If AI systems begin producing conclusions that are functionally correct but cognitively opaque, traditional notions of explanation and accountability may fracture.
That development could alter:
-
Judicial review
-
Regulatory enforcement
-
Contract interpretation
-
The distribution of legal expertise
In this framing, the AI debate becomes less about automation and more about institutional design.
Practical Implications for the Legal Profession
For lawyers and legal institutions, the episode offers several forward-looking implications:
-
Power Shifts, Not Just Productivity Gains: AI changes who can navigate legal complexity effectively.
-
New Forms of Legal Strategy: Models can test hypotheticals, probe vulnerabilities, and simulate regulatory outcomes before disputes arise.
-
Legitimacy Risks: Overreliance on opaque systems may undermine trust in legal decision-making.
-
Professional Adaptation: Lawyers must understand not only what tools can do, but how they reshape responsibility and authority.
The episode ultimately argues that AI is not simply another wave of legal technology. It is a structural intervention into the architecture of legal reasoning itself. Whether that intervention democratizes access to justice or deepens asymmetries depends on how the profession responds.
Additional Resources
New Law Order: Listen to the podcast co-hosted by Joel Cohen and Yale Law Professor John Morley on the current shake-up in the legal services world. Listen on Apple Podcasts or Spotify.
