Gillian K. Hadfield named Bloomberg Distinguished Professor of AI Alignment and Governance

In a world where machines and humans are increasingly intertwined, Gillian Hadfield is focused on ensuring that artificial intelligence follows the norms that make human societies thrive.

"The fundamental thing I'm thinking about is the 'AI alignment problem': how to make sure that AI systems behave in ways that are good for us as humans, in ways that don't disrupt our complex societal structures such as our economies and our political systems, and that these systems do things that we want them to and follow human norms," Hadfield says.

Hadfield, who recently joined Johns Hopkins University as the Bloomberg Distinguished Professor of AI Alignment and Governance, is recognized internationally for her pioneering research in technology, law, and institutional economics. As a faculty member of the university's School of Government and Policy and the Department of Computer Science in the Whiting School of Engineering, Hadfield leads research to reimagine how systems can evolve to meet the demands of a changing world and emerging technologies.

About the BDP

• Name: Gillian K. Hadfield
• Title: Bloomberg Distinguished Professor of AI Alignment and Governance
• Appointments: School of Government and Policy; Department of Computer Science, Whiting School of Engineering
• BDP cluster: Promoting and Governing Technological Advances
• Previous role: Schwartz Reisman Institute for Technology and Society, University of Toronto
• Education: BA, Queen's University; JD, Stanford Law School; PhD in Economics, Stanford University

"The world is moving very rapidly to build increasingly powerful AI agents, and AI is becoming increasingly autonomous, making decisions and having impacts on the world," Hadfield says.

Hadfield approaches the AI alignment problem both in a technical sense, thinking about how best to build AI agents, and from a policy perspective, thinking about legal and regulatory structures needed to guide AI in valuable and not destructive directions. Her approach is grounded in a career thinking about human norms and institutions that promote human well-being.

Hadfield believes that in order for AI to be beneficial to society, AI agents must be built to understand and respond to human normative systems, including both informal norms and formal systems of law. Motivated by a desire to introduce new ways of thinking about what alignment might mean, Hadfield works on computational models of human normative systems to ultimately enable AI systems that align with human normative institutions and reasoning. She cautions that the rapid introduction of AI agents that do not behave the way humans would, on a large scale and in our complex economies, could cause system-wide disruption.

Hadfield's work challenges conventional thinking about how legal rules are made and enforced, and she has become a leading voice in calls to redesign legal infrastructure to serve a globalized and digitally transformed society. Hadfield says that her focus on the AI alignment problem builds on a topic she has spent much of her career as an economist and legal scholar thinking about: incomplete contracting.

"You can hire someone to do a job, but you can never perfectly explain what it is you want them to do," Hadfield explains. "There are always going to be things that are unexpected. When you hire a person to do a job and unexpected things happen, you can have some, albeit not complete, confidence that they're going to behave in ways that are consistent with what you would have wanted, because they know about the norms and laws of society. When we train our AI systems, they currently don't have that. We don't realize how much we take for granted when we delegate tasks or work to someone else; we just assume that a certain structure is going to be there."

Hadfield hopes to contribute the kind of forward-thinking leadership that is critical at this turning point in history, as artificial intelligence rapidly reshapes many aspects of society.

"I'm inspired by the urgency of the AI transformation that we're experiencing," Hadfield says. "What's happening is so fundamentally transformational that we can't use our old approaches to governance and policy and law. We need truly innovative thinking. I want to be thinking about these big questions, such as 'What is justice?' and 'How do we design societies for good?' and I want to do that in ways that are formal and structured. We're at a very important point in history, and I want to be part of the team helping to make sure this goes well."

Hadfield comes to Johns Hopkins from the University of Toronto as part of the Promoting and Governing Technological Advances BDP Cluster. Hadfield says the BDP program and the cluster framework support the interdisciplinarity required for the technical and theoretical innovation needed.

"Coming to Johns Hopkins gives me enormous potential for collaboration," Hadfield says. "Fundamentally, we need to learn to speak across different disciplines, and to be interested in other disciplines. We need lessons and expertise from social sciences brought to bear on how we're doing technical work. The great opportunity is to have highly motivated and impactful researchers who are thinking about the same issues and who are pulled together around shared questions and shared commitments to good answers to those questions."

As a Bloomberg Distinguished Professor, Hadfield joins an interdisciplinary cohort of scholars working to address major world problems and teach the next generation. The program is backed by support from Bloomberg Philanthropies.

"Gillian Hadfield's groundbreaking research on emerging technologies makes her a singular voice in the global conversation on AI and governance," says William Howell, the inaugural dean of the School of Government and Policy. "With new technologies presenting extraordinary challenges and opportunities for democratic institutions, we are thrilled to welcome her to Johns Hopkins."

Adds Ed Schlesinger, dean of the Whiting School of Engineering: "Professor Hadfield brings an extraordinary interdisciplinary vision to Johns Hopkins, advancing the frontier of AI alignment by bridging technical innovation with legal and economic insight to ensure that artificial intelligence serves—and strengthens—human society."