Professor Margaret Hu Joins the Washington Signal Lunch Hour Podcast to Discuss her book, AI Law and Policy

In a wide-ranging conversation on Washington Signal’s Lunch Hour podcast, Professor Margaret Hu discussed the impact of artificial intelligence (AI) systems on constitutional rights, as discussed in her book, AI Law and Policy (Aspen Publishing).

Margaret Hu, the Davison M. Douglas Professor of Law and Director of the Digital Democracy Lab at William & Mary Law School, appeared on the Washington Signal’s Lunch Hour podcast, hosted by Andrew Langer, as part of a series examining artificial intelligence (AI) and the law. The conversation ranged from discussing her career at the Department of Justice and scholarship as a law professor, including authoring AI Law and Policy – a leading casebook and textbook on AI and the law. 

Through the conversation, she emphasized a central concern spanning her career: that opaque, privatized, and largely unregulated algorithms and AI systems are outpacing the constitutional structures designed to protect individual rights.

Professor Hu began her career at the U.S. Department of Justice, Civil Rights Division, the day before the September 11 attacks. From that vantage point, she watched the rapid expansion of predictive analytics and government data systems in the years that followed. She grew concerned that these tools were straining constitutional freedoms, as well as the impact of these systems on civil rights and liberties. 

Central to Professor Hu's argument is that "black box" algorithms and AI systems threaten due process by making it difficult, if not impossible, for individuals to challenge decisions made about them. When automated systems assess risks, flag suspects, or otherwise inform enforcement decisions, Hu explained, the opacity of the underlying algorithm and automated decisionmaking can undermine a person's ability to understand, and confront and appeal, the evidence and reasoning used by the government to take action. 

Hu argued that algorithmic discrimination poses an equally serious threat to equal protection, but one that existing legal doctrine is poorly equipped to address. Traditional discrimination law asks whether a person belongs to a protected class that was treated differently than others. AI systems complicate that framework: a system can be applied uniformly across an entire population and still produce disparate outcomes because of subtle biases buried in its programming. Hu described this as a kind of inversion of the old "separate but equal" doctrine — an "equal but separate" problem, in which formally identical treatment conceals unequal results, without giving courts and plaintiffs a clear doctrinal hook or legally cognizable harm to intervene.

The conversation also turned to the First Amendment, where Hu pointed to bulk metadata collection and cybersurveillance as threats to associational and expressive freedoms. When government, or private companies working on its behalf, can assemble a granular picture of a person's associations and movements, the resulting risk and threat assessments can chill freedom of assembly, association, and the right to petition — even when no single piece of collected data would independently raise concern.

Running through all of this was what Hu described as the deepest structural danger: the privatization of government power. Much of the infrastructure driving AI-based decision-making is built and operated by private contractors rather than government agencies themselves, an arrangement that can blur constitutional accountability. She then floated her idea of a “separation of AI church and state" to better protect people by preventing a single institution from consolidating too much power. 

Asked how the country can come together to regulate AI and curb the powerful interests it has created given today’s deep political division, Hu drew on America's revolutionary history. More specifically, she considered Thomas Jefferson’s rationale when he founded William & Mary Law School in 1779 in the middle of the Revolutionary War. Jefferson understood, she said, that a durable rule-of-law society required not just founding documents but the systemic education of citizens equipped to defend their own rights. The same foresight is needed today, as the country marks the 250th anniversary of the Declaration of Independence.

The full conversation on the Washington Signal’s Lunch Hour podcast is available here.