Jiaee Cheong
I develop methods to evaluate, assure, and improve AI systems that learn, adapt, and act in interaction with people and their environments over time, with the goal of building trustworthy adaptive AI that reliably supports human flourishing.
Jiaee Cheong is a Research Fellow working across Harvard University, MIT, and the University of Cambridge, and a contributor to the United Nations Independent International Scientific Panel on AI. She holds a BSc in Mathematics and Statistics from Imperial College London and a Ph.D. in Machine Learning from Cambridge. Her doctoral research received the 2025 AAAC Outstanding PhD Dissertation Award, awarded annually to a single recipient, and she received the 2024 Cambridge Philosophical Society Research Excellence Award.
Her research focuses on developing trustworthy and adaptive AI systems for human flourishing. She studies how increasingly capable AI systems can remain reliable, safe, and aligned with human needs as they learn and evolve, developing methods for AI evaluation and assurance, metacognitive and self-evolving agents, and human-centred AI. Her work spans AI safety, foundation models, human-AI interaction, and health and wellbeing. She has been supported by several prestigious fellowships and awards from Harvard's Berkman Klein Center, DAAD, the Leverhulme Trust, the Cambridge Trust, and the Alan Turing Institute.
Jiaee Cheong is a Research Fellow working across Harvard University, MIT, and the University of Cambridge. She also contributes to the United Nations Independent International Scientific Panel on AI by supporting the development of its scientific assessments on artificial intelligence. She holds a BSc in Mathematics and Statistics from Imperial College London and a Ph.D. in Machine Learning from the University of Cambridge. Her doctoral research received the 2025 AAAC Outstanding PhD Dissertation Award, awarded annually to a single recipient.
Her research is concerned with developing trustworthy and adaptive AI systems that advance human flourishing. A central question underlying this goal is how AI systems can remain reliable, safe, and aligned with human needs as they learn, adapt, and operate in increasingly complex real-world environments. She approaches this question by developing methods for evaluating and assuring trustworthy AI, studying metacognitive and self-evolving AI agents, and designing human-centred AI systems that can reason about uncertainty and adapt responsibly over time. Her work spans AI safety and reliability, multimodal and foundation models, human-AI interaction, and AI for health and wellbeing.
Jiaee has been supported by prestigious fellowships and awards including the Harvard BKC Fellowship, the Cambridge Philosophical Society Research Excellence Award, DAAD AINet Fellowship, the Leverhulme Trust and the Alan Turing Institute's Doctoral studentship. Across her work, she collaborates with researchers across computer science, medicine, and the social sciences, with the broader aim of building AI systems whose increasing capabilities are accompanied by corresponding advances in trustworthiness, adaptability, and benefit to people and society.
Jiaee Cheong is a Research Fellow working across Harvard University, MIT, and the University of Cambridge, and a contributor to the UN Independent International Scientific Panel on AI. Her research advances trustworthy and adaptive AI for human flourishing, focusing on AI evaluation and assurance, metacognitive and self-evolving agents, and human-centred AI. She studies how increasingly capable AI systems can remain reliable, safe, and aligned with human needs as they learn and evolve. She received her Ph.D. in Machine Learning from Cambridge, where her doctoral research received the 2025 AAAC Outstanding PhD Dissertation Award. She has also received the 2024 Cambridge Philosophical Society Research Excellence Award and fellowships and support from Harvard, DAAD, the Leverhulme Trust, and the Alan Turing Institute.
Opportunities. I am currently seeking motivated individuals for a few ongoing research projects. Please fill in the interest form with your background, interests, and availability.
Research interests
Trustworthy and adaptive human-centred AI
I study how increasingly capable AI systems can remain reliable, safe, and aligned with human needs as they learn and evolve. My work develops methods for AI evaluation and assurance, metacognitive and self-evolving agents, and human-centred AI, with applications spanning AI safety, foundation models, human-AI interaction, and health and wellbeing.
Experience and education
-
Undergraduate
BSc in Mathematics and Statistics
-
Doctoral research, 2025 AAAC award
Ph.D. in Machine Learning
-
Current
Consultant, Independent International Scientific Panel on AI
-
Current
Harvard University , MIT , University of Cambridge
Research Fellow working across Harvard, MIT and University of Cambridge
Selected publications
Full list on Google Scholar* denotes joint-first authorship; † denotes corresponding author; # denotes shared last authorship.
1. Foundations of Trustworthy & Responsible AI
Developing principled methods to evaluate, measure, and assure the reliability, safety, fairness, and robustness of AI systems.
-
KINDER: Kernel-based Independence for Fair Representation Learning via Prototype-space Erasure · NeurIPS, 2026 -
FairSSL: Fair Multimodal Self-Supervised Learning · ICML, 2026 -
Uncertainty as a fairness measure · JAIR, 2024
More publications (11)
- Machine learning fairness for depression detection using EEG data · IEEE ISBI, 2025
- FairGRPO: Towards Fair Reasoning Foundation Models for Clinical Diagnosis · GenAI for Health Workshop, 2025
- Causal fairness for affect recognition · NeurIPS AFCP Workshop, 2023
- The hitchhiker’s guide to bias and fairness in facial affective signal processing: Overview and techniques · IEEE Signal Processing Magazine, 2021
2. Fundamentals of Adaptive & Agentic Intelligence
Designing learning and reasoning methods that enable AI systems to adapt, self-improve, and act autonomously while remaining reliable and aligned.
-
QoQ-Med3: A multimodal reasoning foundation model for clinical analysis · npj Digital Medicine, 2026 -
-
Towards causal replay for knowledge rehearsal in continual learning · AAAI Bridge Program on Continual Causality, 2023
More publications (2)
- Mindbench.ai: An actionable platform to evaluate the profile and performance of large language models in a mental healthcare context · NPP—Digital Psychiatry and Neuroscience, 2025
3. Human-Centred AI
Building interactive and multimodal AI systems that understand, collaborate with, and adapt to people and society, with applications in health and wellbeing.
-
Small but Fair! Fairness for Multimodal Human-Human and Robot-Human Mental Wellbeing Coaching · “It’s not Fair!” – Fairness for a Small Dataset of Multi-modal Dyadic Mental Well-being Coaching · IEEE Transactions on Affective Computing, 2025; ACII (Best Paper), 2023 -
-
Towards gender fairness for mental health prediction · IJCAI (Oral), 2023
More publications (10)
- Equitable Robotics for Wellbeing (Eq-RW) · ACM/IEEE HRI, 2026
- The Current State/Trends in Digital Phenotyping for Mental Health Research and Care · Psychiatric Clinics, 2026
- H3-MOSAIC: A multimodal generative AI model for semantic place detection from high-frequency GPS on H3 grids in mental health geomatics · International Journal of Health Geographics, 2026
- Mobile cognitive remote assessment of schizophrenia: a global multi-site pilot study · Schizophrenia, 2025
- Exploring Causality for HRI: A Case Study on Robotic Mental Well-being Coaching · IEEE RO-MAN, 2025
- Interpreting psychiatric digital phenotyping data with large language models: a preliminary analysis · BMJ Mental Health, 2025
- Causal-HRI: Causal Learning for Human-Robot Interaction · ACM/IEEE HRI, 2024
Honours, fellowships, and grants
- 2026
- 2026
- 2025
- 2025
- 2024
- 2024
Cambridge Philosophical Society Research Excellence Award
- 2024
- 2023
- 2023
- 2023
Peter Clarke Prize for Scientific Communication
Show all awards (~$475k)
- 2026
Harvard Berkman Klein Center Fellowship
Award value: $75,000 annually.
- 2026
- 2022–May 2025
Leverhulme Centre for the Future of Intelligence (LCFI) Fellow
- Finalist, 2025
Imminent Research Grant
- 2025
ISBI Travel Grant
Award value: $900.
- 2025
AAAC Outstanding PhD Dissertation Award
One global award per year. Award value: $1,000.
- 2025
Deutscher Akademischer Austauschdienst (DAAD) AInet Fellowship
Award value: €20,000.
- 2024
University of Oxford MPLS RisingWise Scholar & Panel Host
- 2024
ML4H Travel Grant
Award value: $800.
- 2024 & 2023
Cambridge Philosophical Society Travel Grant
Award values: £1,000 and £500, respectively.
- 2024
Deutscher Akademischer Austauschdienst (DAAD) AInet Fellowship
Award value: £2,600.
- 2024
Cambridge Philosophical Society Research Excellence Award
Award value: £5,000.
- 2024
- 2023
Awarded to pursue research abroad in recognition of research excellence. Award value: £35,000; 15% acceptance rate.
- 2023
Best Paper in Responsible Affective Computing
Awarded at the IEEE International Conference on Affective Computing and Intelligent Interaction (ACII), recognising the top 1% of total submissions.
- 2023
Peter Clarke Prize for Scientific Communication
1st prize winner out of 60+ submissions, awarded for excellent scientific communication writing. Award value: £1,000.
- 2022
Pembroke College Travel Grant
Award value: £1,500.
- 2022
NeurIPS Travel Grant
Award value: £1,500.
- 2020
The Alan Turing Institute Doctoral Studentship
Fully funded PhD studentship. Award value: £125,000; awarded to 30 candidates from 300+ applications.
- 2020
The Cambridge Trust Scholarship
PhD studentship covering tuition fees. Award value: £45,000.
- 2018
Dean’s List
Top in Class for MSc studies at Warwick University.
- 2014
Axiata Young Talent Scholarship Award
Full scholarship covering international tuition fees and cost of living. Award value: £140,000; awarded to 4 candidates from 200+ applicants.
For research collaborations, speaking opportunities, or other enquiries, get in touch by email .
Opportunities. I am currently seeking motivated individuals for a few ongoing research projects. Please fill in the interest form with your background, interests, and availability.