Trustworthy Artificial Intelligence Laboratory
University of Cambridge
We build TRustworthy Artificial intelligenCE systems for high-stakes settings like healthcare, education, and public safety. Drawing from machine learning, human-computer interaction, and social science, we investigate when and how AI can be trusted, integrated into decision-making, and evaluated in ecologically valid settings.
Research
Deployment as a Science
Systematic approaches to document how AI systems are deployed across diverse real-world contexts.
Human-AI Externalities
Investigating the broader impacts of AI on individuals and society—both intended and unintended.
Agent Orchestration
Frameworks for deploying AI agents that work alongside humans in complex environments.
Trustworthy AI
Building AI systems that are reliable, fair, transparent, and aligned with human values.
Team
Umang Bhatt
PI, Assistant Professor
Cambridge · CHIA
Elaf Almahmoud
PhD Student
Cambridge
Noah Broestl
PhD Student
Cambridge
Drew Calcagno
PhD Student
Cambridge
Bernardo Villegas
PhD Student
Cambridge
Yung-Hsuan Wu
Research Assistant
Cambridge
Olivia Slack
Lab Manager
Cambridge
Kostadin Devedzhiev
Masters Student
Cambridge
Afan Handoyo
Masters Student
Cambridge
Shin Zert Phua
Masters Student
Cambridge
Juan Pablo Ruiz
Masters Student
Cambridge
Ariella Shulman
Masters Student
Cambridge
Joy Wang
Masters Student
Cambridge
Affiliated Students
Friends of TRACE
Recent Publications
Towards Interactive Evaluations for Interaction Harms in Human-AI Systems
A framework for evaluating interaction harms in human-AI systems through interactive, dynamic assessment methods.
arXiv 2024
2024Building Machines that Think with People
A vision for human-AI collaboration leveraging complementary strengths of cognition and machine learning.
Nature Human Behavior
2025Personalized Decision Support Policies
Learning decision support policies that adapt to individual users, improving team performance.
AAAI 2025
News
Umang taught a three-week course on Responsible AI at the African Institute for Mathematical Sciences (AIMS) in Cape Town, South Africa. Three AIMS students — Hassan Suliman, Dagmawi Misker, and Ghaida Ibrahim — are now joining the lab as affiliated researchers. Read more on LinkedIn.
The hidden costs of agentic AI: humans recruited to sense the physical world on an agent's behalf, potentially without their consent. We're all sensors. Umang explores this in Noema Magazine and on Substack.
Umang and Ariella will represent the University of Cambridge at the India AI Impact Summit 2026.
Our Supporters
Industry Partners
Global Academic Partners
