AI & GOVERNANCE
Artificial Intelligence and Law
Exploring algorithm explainability, allocation of responsibility, and institutions for fair decision-making.
Perspective of This Topic
Enhancing AI performance and socially trustworthy use require different analyses. This research agenda examines at each stage—design, training, deployment, and use—who can identify and control risks. Attention is paid not only to the content of rules but also to procedures that enable rules to function effectively, such as audits, objections, and remedies.
Three Axes of Analysis
- Allocation of Responsibility: comparing the information and control capabilities of developers, adopting institutions, and users.
- Conditions for Verification: distinguishing what performance metrics, explainability, and independent evaluations demonstrate.
- Comparison of Policy Instruments: reading together the costs and effects of ex ante obligations, ex post liabilities, standards, and certifications.
Related Projects and Research Materials
Connected Research Commentary
To Whom Should Algorithms Be Accountable?
Considering Explainability and Responsibility Together