by Beth Hester
Among her many professional achievements, Hu is a Faculty Affiliate with the Global Research Institute and Data Science at W&M, and a Research Affiliate with Pennsylvania State University’s Institute for Computational and Data Sciences. Her research focuses on the intersection of civil rights, national security, cybersurveillance, and AI. She is author of several notable works, and previously served as Special Policy Counsel in the Civil Rights Division of the U.S. Department of Justice.
CoVaBIZ: I’ll start with a rather long question, but I think it’s necessary to set the stage. What gap in legal education or practice did you observe that prompted you to create an AI-focused casebook? I know you have a particular focus on the law and intersectionality especially when it comes to data privacy and surveillance, AI and power structures, and the impacts of algorithmic decision-making on constitutional democracy. That’s a lot. So how did your background as a legal scholar inform the way the volume took shape?
Hu: Around the academic year of 2017-18, Aspen Publishing first approached me and explained that there was a lack of curricular tools for AI and the Law courses at law schools. They started to see a growing trend among law schools that were starting to offer seminars on AI in course catalogues, however, there were no textbooks in the field. The courses covered AI-related legal topics such as cybersurveillance and algorithmic discrimination, and the constitutional dimension of some of the technological developments surrounding AI systems and automated decision-making. My published articles covered those types of topics. So, the editors at Aspen explained that they thought that I should start the process of writing this type of casebook or textbook.
Before I became a law professor, I previously served in the federal government after the terrorist attacks of September 11, 2001. I started in the Civil Rights Division of the U.S. Department of Justice the day before the attack, on September 10, 2001, and served in the Division for almost a decade. I provided policy guidance on data-driven systems and databasing and data analytics programs supporting immigration enforcement and homeland security goals. My AI research draws upon my work in the U.S. Department of Justice and U.S. Department of Homeland Security.
How do you see courts currently struggling with evolving AI-related litigation? Do you have an opinion as to the kind of AI-related disputes most likely to reach the Supreme Court in the not- too-distant future?
State and federal judges are now struggling daily to keep up with a flood of litigation involving a wide range of issues that involve AI technologies and algorithms, such as social media, generative AI, automated decision-making systems, facial recognition AI, and other systems, often that impact our rights.
AI video generators and text-to-video AI systems, for example, are challenging traditional intellectual property laws in unprecedented ways. The platform economy and deep fakes and how it is regulated by either the government or corporations can profoundly impact our First Amendment. AI-driven sensors that are now embedded in our vehicles and medical diagnostics that depend upon AI implicate tort and contract law. So, all of these types of disputes, most likely, will be journeying to the Supreme Court.
What types of legal practitioners will most need fluency in AI law over the next decade and do you expect AI literacy to become as essential as constitutional or contracts law?
My prediction is that everyone who engages with the law will need to be fluent in AI law and policy. If we have a constitutional democracy that is built on evidence-based reasoning, then a common understanding of truth and objective reality is critical, and many AI systems can challenge that. The reinvention of reality through AI directly conflicts with verifiable truth. An inability to discern reliable evidence and legal research challenges a rule of law governance structure. Contracts and other legal instruments, including laws themselves, that are drafted by AI tools will be increasingly common and will need to be scrutinized carefully for errors, bias, and other issues.
Where do you see the greatest tension between innovation and regulation in AI?
Although some argue that regulation defeats AI innovation, I think the opposite is true: AI innovation demands strong legal regimes. For example, intellectual property laws incentivize inventions and experimentation of ideas that can be protected and monetized. This has required strong institutions that can enforce these legal regimes.
What unresolved legal question around AI keeps you up at night?
So many! One legal question that is increasingly raised is whether AI-driven robots or chatbots are persons under the law and should be recognized with personhood rights legally, or even citizenship rights. That type of theory of rights can raise a real Pandora’s box for lawyers, lawmakers, and judges.
If you were advising policymakers today, what principle should guide AI regulation?
My book, AI Law and Policy (Aspen Publishing), starts with the concept of rights and principles of constitutional law and fundamental rights. Then, as a pedagogical theory, I invite teachers and students, and anyone reading the book outside of the classroom, to think about whether the technological or legal developments are consistent or not consistent with the rule of law and the ideals that anchored our constitutional republic at our founding and reconstruction. That’s the primary principle that I think should guide policymakers.
It seems that regulation of AI is still largely fragmented, and there are wildly divergent levels of tech literacy among lawmakers, but are you starting to see the emergence of a coherent federal AI regulatory or oversight framework?
Sadly, no. There is no comprehensive federal regulatory or statutory framework for AI oversight at present. Additionally, in the U.S., we currently lacking an omnibus oversight framework for cybersecurity or information security, and data privacy or data protection. All of these areas are deeply interrelated: AI, information security, and data privacy. Other nations are racing ahead in legal innovation. In the U.S., I hope that we realize that national security strategies should engage in technological and legal innovation to address this rapidly evolving landscape.
In an ideal world what would a “successful” AI regulatory system look like from your perspective? Do you think it’s achievable in the near future?
Over the past decades, we’ve started to see some meaningful experimentation in federal, state, and local governments in the U.S. We’ve seen public-private partnerships and multi-stakeholder engagement, such as the collaborative efforts by the U.S. Department of Commerce’s NIST experts who have been building out the AI Risk Management Frameworks.
Companies have been adopting AI Ethical Frameworks. External auditors have been building testing and AI sandboxes. And many nations outside of the U.S. and the EU have started experimenting with AI regulatory frameworks. Judges are starting to grapple deeply with these issues, and we’re witnessing an emerging jurisprudence. I think we need to keep our eyes on the birth of this field and, like the evolution of other areas of law over the past few decades, such as environmental law, it is a matter of trial and error.
Now a personal question: If you could invite three people, living or dead, to a dinner party with family and colleagues, who would they be and why?
Because it is the 250th birthday of the nation this year, I would love to dine with Thomas Jefferson and George Wythe, his lawyer mentor who took Thomas in as an apprentice, to discuss the intellectual foundations of the nation and the philosophical anchoring of the Declaration of Independence, in particular. And then I would love to invite Annette Gordon-Reed, the Harvard Law professor who has dedicated the past few decades to studying Jefferson and the Hemings family and their lives together in Monticello, to help curate the questions.
I would like to structure the dinner party as a traditional “Jefferson Dinner,” a small salon-type dinner of around 10 that Jefferson was famous for hosting to discuss the most pressing and exciting questions of the day. I would like to present the dinner guests—a combination of distinguished leaders from industry, government, civil society, and the academy—a series of provocation questions to be on the theme of how to set the conditions of success for the democratic republic for the next 250 years.
