Regulatory Consulting at Government Performance Lab (Harvard)

Overview

The Harvard Kennedy School Government Performance Lab (GPL) is an applied research lab based in HKS’s Taubman Center for State and Local Government, focused on helping governments redesign social-service systems — child and family well-being, safety and justice, homelessness and housing — around outcomes rather than outputs. Its core method is not AI-specific: the GPL helps government leaders identify challenges, design solutions, test and demonstrate effectiveness (often through outcomes-based contracting), and scale what works. The lab is best known as one of the founding technical-assistance partners of What Works Cities, the Bloomberg Philanthropies initiative that has certified more than 100 cities worldwide for data-driven governance, alongside partners including the Center for Government Excellence (GovEx) at Johns Hopkins, Results for America, and the Behavioral Insights Team. As AI becomes a more prominent tool inside the social-service systems the GPL works on, its outcomes-based discipline is a natural, though still emerging, complement to the broader AI-in-government conversation happening elsewhere at Harvard Kennedy School.

AI-Focused Regulatory Work

The honest way to place the GPL in an AI-policy directory is as an outcomes-and-implementation specialist whose core discipline — rigorously testing whether a government intervention actually works before scaling it — is exactly the kind of evaluative rigor that responsible AI adoption in government will require, even though the lab’s own publication record is not yet centered on AI specifically. That AI-adjacent context is developing quickly around it: Harvard Kennedy School announced a ten-year strategic blueprint in 2026 that includes a new tech-policy concentration for its Master in Public Policy students and the hiring of AI-and-policy faculty, including MIT’s Stephen Casper to teach AI governance; HKS lecturer Mark Fagan co-authored ‘Governing With AI: How the Public Sector Can Use Artificial Intelligence to Improve Performance’ and a public-sector AI risk-assessment framework arguing that agencies should weigh the costs of AI’s risks against the real costs of delaying its benefits rather than defaulting to a precautionary block. As HKS Dean Jeremy Weinstein put it, ‘there is no policy area, there is no function in the public or social sector, that isn’t going to be massively impacted by the accelerating power of generative AI’ — a mission GPL’s outcomes-based methodology is well positioned to serve as governments begin asking not just whether to adopt AI in social services, but whether it demonstrably improves the outcomes GPL specializes in measuring. For a directory reader, the GPL is most valuable as a rigorous-evaluation partner and a strong networking node inside HKS’s fast-growing AI-and-public-policy ecosystem, rather than as a source of AI-specific publications today.

Notable AI Policy Work

• Co-founded and serves as a lead technical-assistance partner of What Works Cities (Bloomberg Philanthropies), which has certified 100+ cities across 10 countries for data-driven governance — a network increasingly focused on how cities use data and AI responsibly.
• Applies outcomes-based contracting and rigorous evaluation methods to redesign government social-service systems in child and family well-being, safety and justice, and homelessness and housing.
• Operates within a Harvard Kennedy School AI ecosystem that in 2026 launched a new tech-policy MPP concentration and hired dedicated AI-and-policy faculty (including MIT’s Stephen Casper) as part of a 10-year strategic blueprint.
• HKS lecturer Mark Fagan (affiliated with the same Taubman Center as GPL) co-authored ‘Governing With AI: How the Public Sector Can Use Artificial Intelligence to Improve Performance’ and a public-sector AI risk-assessment framework.
• Helps government leaders identify challenges, design and test solutions, and scale promising practices — a methodology directly applicable to evaluating whether AI pilots in government actually improve resident outcomes.

Thought Leadership

1. Harvard Kennedy School Government Performance Lab — About Us
   The GPL’s mission and outcomes-based methodology for redesigning government social-service systems.
2. AI for Governments and Policymakers (HKS)
   HKS’s broader AI-and-public-policy research hub, including lecturer Mark Fagan’s ‘Governing With AI’ book and public-sector AI risk-assessment framework.
3. How AI May Change Public Policy, How We Lead (Harvard Gazette)
   Coverage of HKS’s 10-year strategic blueprint, including a new tech-policy MPP concentration and AI-focused faculty hires, that now surrounds the GPL’s work.
4. What Works Cities: What Excellence Looks Like in Local Government
   Documents GPL’s role as a founding technical-assistance partner in the Bloomberg Philanthropies What Works Cities initiative.

Clients

State and local governments across the U.S. and, via What Works Cities, internationally (100+ certified cities in 10 countries). Key partners include Bloomberg Philanthropies, Results for America, the Center for Government Excellence (GovEx) at Johns Hopkins, and the Behavioral Insights Team.