Hugh & Hazel Darling Foundation Professor at the University of San Diego School of Law, and Director of the Center for the Study of Constitutional Originalism

Joined June 2009
Michael Rappaport retweeted
Tune in now at vthepeople.org for our first panel of the day on the history and traditions of Article V!
6
4
2
297
Is there any government action regarding AI that might be beneficial? Helen Toner argues that we need a National Transportation Safety Board-type entity to investigate AI incidents. Significantly, the agency would not have regulatory authority. nytimes.com/2026/09/21/opini…
1
1
393
This seems like a promising approach.
1
183
Michael Rappaport retweeted
Recently, calls for an AI agency and AI regulation have become louder, including calls from the AI companies themselves. (E.g. @CassSunstein) The AI company calls raise the question whether future AI regulation will be public interested or will further the interests of these
1
1
1
3
415
Recently, calls for an AI agency and AI regulation have become louder, including calls from the AI companies themselves. (E.g. @CassSunstein) The AI company calls raise the question whether future AI regulation will be public interested or will further the interests of these
1
1
1
3
415
companies at the public's expense. Are these calls by the companies really about safety or about corporate protection? Here is one thing to look for. When the regulation comes, will it insulate the companies from liability for problematic behavior from their products? If so, then there is a strong argument that the regulation is not promoting safety but is about protecting the companies. Liability for AI companies is one of the strongest ways of providing the companies with the incentive to be prudent and safe. I asked Gemini about the reasons, other than safety, that the AI companies might favor regulation. Here is what it said: Even if AI companies completely set aside safety, ethics, and existential risk, they still have massive business incentives to lobby for and support AI regulation. In corporate strategy, regulation is often used as a competitive tool. Here is why AI companies would actively push for regulation from a pure business perspective: 1. Regulatory Capture and Moats (Protecting Market Incumbents) Raising Compliance Costs: Comprehensive regulations—such as mandatory multi-million-dollar safety audits, third-party evaluations, and strict reporting requirements—are minor line items for well-capitalized tech giants. However, these same costs can be fatal to early-stage startups and open-source developers. Licensing Requirements: If governments require official licenses to train frontier models above a certain computational threshold , existing leaders effectively secure an oligopoly, locking out future competitors before they can scale. 2. Legal Immunity and Standardized Liability Rules Safe Harbors: Unclear legal frameworks leave companies exposed to unpredictable, high-stakes lawsuits over issues like copyright infringement, data privacy, and defamation. Shifting Responsibility: Government-backed standards create a legal defense shield. If a company complies with official regulatory benchmarks, courts are far less likely to hold them liable when a system malfunctions or causes commercial harm. 3. Market Stabilization and Consumer Trust Enterprise Adoption: Risk-averse enterprise clients (such as healthcare providers, banks, and government agencies) are often hesitant to adopt unregulated, legally dubious software. Official government standards legitimize AI tools and unlock high-margin enterprise contracts. Preventing Race-to-the-Bottom Volatility: Regulation provides a predictable framework. Without clear ground rules, reckless market entrants might ship unreliable products that trigger severe public or economic backlash, harming the entire sector's customer confidence. 4. Fragmented Law Avoidance (Preemption) National vs. Local Rules: Lacking federal regulation, individual US states and international regions create a chaotic patchwork of conflicting laws (e.g., varying state-level deepfake or transparency statutes). Single Standard Preference: AI companies prefer a single, unified federal or international regulatory framework over negotiating thousands of different regional compliance standards. Key Takeaway: In corporate history, major industries—from telecommunications and banking to pharmaceuticals—have routinely welcomed regulation. When an industry matures, regulation transitions from an obstacle into a defensive barrier to entry, cementing the dominance of market leaders while stabilizing their business models.
3
265
Cass Sunstein @CassSunstein recommends an agency to regulate AI. But given that agencies often produce harmful regulations that reduce competition and reduce safety, his recommendation would be more persuasive if he addressed the government failure problem.
3
7
1,122
Michael Rappaport retweeted
Replying to @MTSlive
@grok Is this true?
2
1
2
15,225
Michael Rappaport retweeted
I'm excited to be joining the team at the Originalism Blog along with Ryan Snyder. originalismblog.com/two-new-…
2
1
38
3,416
The Constitution’s meaning does not run out in close cases merely because the issue is contested. Originalism can stay interpretive—not normative—all the way down. Read the Law and Liberty Essay by John McGinnis and me on Originalism’s Better Meaning. lawliberty.org/originalisms-…
2
1
7
1,354
NRO on AI: "No industry in the history of the world has marketed itself with such imbecilic self-sabotage: What we’re building is horrible, will ruin your livelihood, and possibly enslave you! Oh, and may we please have a few acres to scale up our operations?
3
4
623
Another interesting essay by @CassSunstein on Hayek. This one on Hayek's essay, Why I am not a Conservative. casssunstein.substack.com/p/…. Sunstein's reference to Hayek's probable view of Buckley's "standing athwart history" quote is excellent.
1
3
1,213
Michael Rappaport retweeted
Interesting question, but not obvious to me what the answer is. I've often looked to 18th century law in my work, and wondered about what formal academic training in history might help me. But my tentative answer is that a lot of the kind of legal history that U.S. legal academics are interested in (in my case, at least, it's often about the legal world, assumptions, and doctrines of the 18th century) is a sort of overlooked intersection between law and history for which it's difficult to get formal academic training.
POLL: When a legal academic purports to be a serious historian, we [should/should not] take note of the fact that he or she has no formal academic training in history.
64%Should
36%Should not
467 votes • Final results
6
8
4
80
39,173
With the right (Vance) criticizing Milton Friedman and the left (Mamdani) criticizing Hayek, it is a sad day for America.
2
3
8
910
The transcribing of the Herculaneum scrolls raises the possibility of discovering lost treasures. I asked Gemini which manuscripts scholars would be most interested in. My own suggestion: Aristotle's lost dialogues. Here is what Gemini said:
1
3
795
The prospect of reading the Herculaneum scrolls via virtual unwrapping is arguably the most thrilling development in modern classical scholarship. For centuries, our window into antiquity has been filtered through medieval copyists, meaning we only have what they chose to preserve (or what physically survived fires, rot, and war). If the Villa of the Papyri—or its still-unexcavated lower levels—yields a substantial Latin or broader Greek library, a few specific "lost works" top the wish list of virtually every classicist and historian. 1. Aristotle’s Dialogues. You hit the nail on the head. The Aristotle we read today (Nicomachean Ethics, Politics, Metaphysics) is famously dense, dry, and structured like lecture notes or internal research treatises. But in antiquity, Aristotle was renowned as a master stylist. Cicero famously described his popular, published works—his dialogues—as a "golden river of eloquence." Works like the Protrepticus (an exhortation to philosophy) or On Justice were written for the public, likely in the beautifully accessible, dramatic style of Plato. Finding an intact copy of these would completely revolutionize our understanding of his philosophical development and literary skill. 2. Claudius’s Etruscan History (Tyrrhenika). The Roman Emperor Claudius was an avid historian who spoke Etruscan. He wrote a massive, 20-volume history of the Etruscans, a deeply influential civilization that predated and shaped early Rome. Because the Etruscan language and culture were entirely swallowed by the Roman Empire, this work is considered the ultimate "holy grail" for understanding early Italian history. 3. The Lost Books of Livy (Ab Urbe Condita). Livy wrote a monumental, 142-book history of Rome from its founding up to his own era (the reign of Augustus). Only 35 of those books have survived to the modern day. We are missing his entire account of the late Roman Republic, including the cataclysmic civil wars, the rise of Julius Caesar, and the collapse of the traditional Roman constitution. 4. Lost Greek Tragedies and Comedies. We only possess a tiny fraction of the works by Athens' greatest playwrights.Aeschylus & Sophocles: We have only 7 surviving plays for each, though they wrote roughly 90 and 120 respectively.Euripides: Only 19 out of 90+ survive.Sappho: Her deeply influential lyric poetry survived almost entirely in scattered fragments and quotes within other texts. 5. Early Roman Literature and Epic Poetry. Works like Ennius's Annales (the foundational epic poem of Roman history before Virgil's Aeneid) exist only in fragments. Similarly, the early Roman comedies, political speeches by figures like Cato the Elder, and Cicero's lost philosophical dialogue Hortensius (which famously inspired St. Augustine to convert to Christianity) are highly sought after. The Reality Check: So far, the scrolls read from the Villa of the Papyri have been overwhelmingly dominated by Greek Epicurean philosophy, particularly the works of Philodemus. This makes sense, as the villa is believed to have belonged to Lucius Calpurnius Piso Caesoninus (Julius Caesar's father-in-law), who was Philodemus's patron. However, scholars remain highly optimistic that a separate Latin library or a deeper main archive remains buried in the unexcavated portions of the estate.
1
512