Algorithmic Fairness and Accountability
Formal methods for auditing automated decision systems, and the policy frameworks that determine whether those audits carry any weight.
About
I work on the mathematics of fairness in automated systems, and on the policy that decides whether any of it matters. I direct Brown’s Center for Tech Responsibility, and I served in the White House Office of Science and Technology Policy, where I helped write the Blueprint for an AI Bill of Rights.
Before returning to Brown I was Assistant Director for Science and Justice at the White House Office of Science and Technology Policy. My research sits at the intersection of algorithmic fairness, computational geometry, and the governance of automated decision systems.
Recent
Testimony before the Senate Subcommittee on Consumer Protection
On algorithmic accountability in consumer lending, and why disclosure alone does not constitute oversight.
The Center for Tech Responsibility launches its audit clinic
A standing partnership pairing Brown students with civil society organizations that need technical review.
New paper accepted to FAccT 2026
On the gap between the fairness metrics we can compute and the harms people actually report.
Appointed to the National AI Advisory Committee
A two-year term, working on the algorithmic accountability subcommittee.
Selected work
Papers
- Fairness and Abstraction in Sociotechnical Systems
- Auditing Automated Decision Systems Without Access to the Model
- On the Limits of Statistical Parity as a Governance Standard
- Certifying Robustness in High-Dimensional Classification
Talks
- What the Blueprint Got Right, and What It Missed
- Auditing Without Access
- Algorithmic Accountability in Practice
- Who Gets to Audit?
Media
- Who Audits the Algorithms?
- On the limits of AI self-regulation
- What an AI Bill of Rights would actually do
- The accountability gap
Research
Governance of Automated Decisions
What regulators can require, what they can verify, and the distance between the two.
Computational Geometry
Structure and approximation in high-dimensional data, the thread running back to the beginning of my work.