A new wave of research is moving beyond abstract AI governance debates to examine how these systems are actually being used in hospitals, courtrooms, and classrooms. Georgetown University's McCourt School of Public Policy announced nine research grants on June 30, 2026, supporting faculty-led projects that investigate AI deployment across criminal justice, healthcare, education, and democratic participation.

Why Are Researchers Mapping AI Use in Public Institutions?

Governments and policymakers are increasingly adopting AI tools, but critical questions remain unanswered: Are they using these systems effectively? What risks do they pose to the people affected by them? The McCourt School's Tech and Public Policy Center is funding research to answer these questions by tracking, learning from, and improving AI deployments in three high-stakes domains.

The research spans three broad themes: understanding how AI systems operate within public institutions, examining public attitudes and online behaviors in an era of algorithmic influence, and evaluating regulatory and governance frameworks for emerging technologies. This approach reflects a shift from theoretical policy discussions to empirical investigation of what's actually happening on the ground.

What Specific AI Systems Are Researchers Investigating?

The funded projects target some of the most consequential uses of AI in American life. In criminal justice, Associate Professor Andrea Headley is leading a comprehensive study that will survey 350 cities and 150 counties to document where and how AI is being deployed across the U.S. criminal justice system. The project will also conduct community-engaged fieldwork to assess how justice personnel and impacted communities perceive these tools' fairness and accountability.

In healthcare, Assistant Research Professor Amy Killelea is investigating how health insurers and third-party data vendors use AI to process prior authorization requests, a widespread and increasingly controversial practice. Her research will assess the consumer risks and benefits of growing reliance on these automated systems.

Internationally, Distinguished Professor Jishnu Das will evaluate whether AI-assisted winter schools can improve foundational learning outcomes among primary school children in Quetta, Balochistan. This project tests whether governments can leverage AI as a scalable model for education in resource-constrained regions.

How Are Healthcare Organizations Preparing for AI Contracts?

As AI adoption accelerates in healthcare, organizations face a complex contracting landscape. Unlike traditional software, AI systems often involve continuous learning, probabilistic outputs, and opaque decision-making processes. These characteristics introduce unique legal and ethical challenges, especially when AI tools inform patient care or handle sensitive health data.

Healthcare organizations should address several critical areas before deploying AI solutions:

• Data Rights and Ownership: Contracts should clearly outline which party owns the underlying data and how the other party may use resulting predictions, insights, or models. If protected health information or de-identified data is used to train or refine algorithms, this raises questions under the Health Insurance Portability and Accountability Act (HIPAA) and state privacy laws.
• Liability and Risk Allocation: AI systems can generate inaccurate or biased outputs that lead to clinical errors, missed diagnoses, or inappropriate care recommendations. Contracts should include robust indemnification provisions and allocate liability for malfunctions, misdiagnoses, or system errors by cause, including model design defects, inaccurate or biased outputs, integration errors, and cybersecurity incidents.
• Transparency and Human Oversight: Contracts should require vendors to notify providers of known defects or risks and establish performance guarantees through service level agreements. These issues are especially relevant when AI tools qualify as software as a medical device under the U.S. Food and Drug Administration's regulatory framework.
• Patient Notice and Privacy Compliance: Patients should receive adequate notice regarding what information is collected, how that information is used or shared, why their information is collected, and who has access to their information. Contracts should include detailed provisions outlining notice requirements to ensure compliance with privacy laws and regulations.

What Should Healthcare Organizations Know Before Signing AI Contracts?

A threshold step in contracting for AI services is to clearly define the scope and purpose of the tool. Agreements should specify whether the AI system is intended to assist in diagnosis, automate administrative tasks, or stratify patient risk. Deliverable requirements should extend beyond implementation and may include ongoing performance expectations, implementation support, testing and monitoring, and other methods for measuring success.

For adaptive AI systems, it is essential to address how updates, retraining, and version control will be managed over time. These details help limit ambiguity and support regulatory compliance, especially when integration with electronic health records or clinical decision support systems is involved.

The contract should identify each party's role under HIPAA, whether as a covered entity, business associate, subcontractor, service provider, or processor. This alignment is critical for data-use permissions, patient request handling, breach notification, and subcontractor flow-down obligations. The agreement should also limit secondary uses of data, prohibit unauthorized commercial exploitation, and require compliance with all applicable privacy regulations.

How Are Researchers Studying AI's Impact on Democratic Participation?

Beyond criminal justice and healthcare, the McCourt School's research agenda addresses how social media, online platforms, and information technology affect society. Assistant Professor Tiago Ventura will lead a team to build a secure, privacy-preserving data-donation pipeline that allows consenting participants to share portions of their conversational AI histories for research. This infrastructure will shed light on how people use tools like ChatGPT in sensitive domains including politics and health, informing governance approaches to reduce harms.

Professor Nejla Asimovic will investigate whether structured, reflective engagement with algorithmic dynamics can reduce outrage-driven reactions and improve how people relate to those in other groups. The project will identify which specific forms of engagement actually shift intergroup attitudes and behavior, and develop scalable tools that put those approaches into practice.

Associate Research Professor Renée DiResta will develop and evaluate whether large language model-assisted counterspeech can help produce clearer, faster, and more evidence-grounded Community Notes to reduce the harms of health misinformation on social media.

What Role Will Public Opinion Play in AI Governance?

Knowing what citizens want is a vital first step in developing effective policy. Associate Professor Jonathan Ladd will conduct a major national survey measuring the public's hopes and fears about the technology industry and AI, and will convene a conference examining the public's evolving relationship with big tech.

"These grants move beyond abstract debates about AI to examine how these technologies are actually being deployed, who they affect, and what guardrails are needed," said Carole Roan Gresenz, Dean of the McCourt School of Public Policy.

Carole Roan Gresenz, Dean of the McCourt School of Public Policy

The research also addresses how U.S. AI regulations shape the adoption of AI abroad. Peter F. Krogh Professor Erik Voeten in the School of Foreign Service will examine how domestic regulatory guardrails could facilitate, rather than hinder, the global diffusion of American AI technologies amid great power competition.

Additionally, Professor Meg Leta Jones and Professor Paul Ohm are working with state attorneys general to generate empirical evidence documenting potential non-compliance with privacy and AI regulations. Their work allows students to produce knowledge that goes directly to policymakers grappling with the complexities of regulating technology.

These research initiatives represent a significant shift in how policymakers approach AI governance. Rather than relying on theoretical frameworks or industry self-regulation, universities are now producing ground-level evidence about where AI is deployed, how it affects communities, and what safeguards are actually needed. As AI adoption accelerates across healthcare, criminal justice, and education, this empirical foundation will become increasingly critical for developing effective and equitable governance frameworks.