Frank Pasquale

In the [YOU] on AI Field Guide

The cycle that began with [YOU] on AI asks what it would mean to take the orange pill—to see the machine clearly, without the narcotic of hype or the paralysis of fear. Pasquale is the clearest seer of the governance gap: the chasm between how these systems perform and how they answer for what they do. He supplies the legal scholar's version of the cycle's central question—not what can a model produce, but to whom is it accountable when it produces something unjust? The cycle situates him as the thinker who refuses the passive posture, the assumption that technology arrives with its social consequences already fixed and that our only choice is to adapt.

His lens reframes every policy question the cycle touches. The issue is never how impressive a system's benchmark score is, but whether the person it acts upon can see the mechanism, name the authority, and contest the decision. Pasquale demonstrates—with legal precision, not rhetoric—that the black box is not a natural feature of complexity but a policy choice: opacity is substantially a product of trade-secret doctrine and nondisclosure regimes that can be changed by different rules. This reframing moves the reader from helplessness to agency, which is the cycle's whole purpose.

He stands in the cycle's gallery alongside thinkers who supply different instruments for the same measurement. Where Shoshana Zuboff names the economic system that harvests human experience as raw material, and algorithmic governance scholars document the progressive replacement of human judgment by automated systems, Pasquale provides the legal-reform blueprint—the specific laws, rights, and institutional structures through which a self-governing society imposes its will on the concentrations of power operating within it.

The deepest alignment with the cycle's spirit is Pasquale's refusal of technological determinism. He regards the rhetoric of inevitability as a political instrument disguised as a technical observation: to say the machines are coming and adaptation is the only response is to help fix the future in the interest of those already building it. The cycle insists, identically, that the future of AI is a human choice rather than a technological fate—and Pasquale has spent his career showing, concretely, what the better choice would require.

Origin

Pasquale trained at Harvard, Oxford, and Yale, clerked for a federal appellate judge, and built his career across Brooklyn Law School, Seton Hall, and the University of Maryland before joining Cornell Tech and Cornell Law School in 2023. The trajectory matters: he arrived at AI not from computer science but from the ancient questions of his discipline—who exercises power, under what constraints, and with what accountability to those they affect. Where the engineer's eye falls on capability, the lawyer's falls on jurisdiction, and Pasquale has spent two decades asking which jurisdiction governs the decisions that algorithms make about human lives.

His 2015 book The Black Box Society mapped three arteries of the economy—reputation, search, and finance—where opaque automated systems had become decisive, and showed that the opacity was not technical but legal: protected by trade-secret doctrine, nondisclosure agreements, and the criminalization of disclosure. The argument that secrecy is a policy choice masquerading as an inevitability became the book's most influential move, because it pointed toward a remedy: different rules. His 2020 New Laws of Robotics extended the project from diagnosis to prescription, proposing four laws—complement professionals, do not counterfeit humanity, avoid zero-sum arms races, always name your creators and controllers—as the enforceable framework for a democratic society's relationship to its machines.

With the legal scholar Danielle Citron, Pasquale developed the concept of technological due process—the insistence that the procedural protections governing consequential decisions, notice of the grounds, the right to contest, the ability to appeal, must accompany decisions wherever they migrate, including into automated systems. This work predated the General Data Protection Regulation's provisions on automated decision-making and remains, in his view, insufficiently implemented anywhere.

Key Ideas

The Black Box Society. Pasquale's central image is a society governed by the double meaning of the black box: the flight recorder that captures everything, and the sealed mechanism that can be described only by its inputs and outputs. We live inside both at once—continuously recorded by systems we cannot read. The asymmetry is the whole of the power: algorithmic opacity is secured not by technical necessity but by trade-secret law, nondisclosure agreements, and the rhetoric of proprietary methods. The black box, he insists, is a policy choice. Different rules can crack it.

The One-Way Mirror. The contemporary world resembles a one-way mirror: platforms can see everything about us and we can see almost nothing about them. To have power over someone is to scrutinize them while remaining unscrutinized yourself. Pasquale draws on the legal scholar Jack Balkin to frame the relation as interrogation without reciprocity—and insists that the remedy is not symmetry of surveillance but symmetry of accountability: those who exercise power over others must be able to justify how they do it.

Four New Laws of Robotics. Taking Asimov as a foil, Pasquale proposes four laws addressed not at machines but at the humans who build and deploy them. AI should complement professionals rather than replace them. It should not counterfeit humanity. It should not intensify zero-sum arms races. It must always indicate its creators, controllers, and owners. The laws describe a world in which complementarity, authenticity, cooperation, and attribution are not voluntary ethics but enforceable public rules.

Complement, Do Not Replace. The first law is the most consequential for how AI is built. Pasquale distinguishes substitutive automation—which aims to remove the human from the payroll—from complementary automation, which aims to make the human more capable. The same underlying technology can be pointed in either direction. He argues that the professions—medicine, law, education, journalism—are not rent-seeking guilds to be disrupted but accumulated repositories of judgment, ethics, and accountable expertise that automation should extend, not replace.

Technological Due Process. Developed with Danielle Citron: as automated systems make consequential decisions about individuals, the basic protections of due process—notice of grounds, access to logic, meaningful opportunity to contest before a human who can be held accountable—must follow the decisions into the machine. Attribution is the precondition: you cannot contest a decision you cannot trace to a responsible party. Pasquale treats the right to an explanation not as a courtesy but as a test that automated systems in high-stakes domains must pass.