About
About the lab
The Human-AI Empowerment (HAIE) Lab was founded in August 2022 in the Human-Centered Computing Division of Clemson University's School of Computing. We study how artificial intelligence can help people achieve their goals, and how to design AI that people can understand, supervise and trust.
Understand how AI-powered tools can empower individuals and communities, and turn that understanding into systems that measurably expand people's opportunities.
Identify the human factors that machines can learn from in order to help us achieve our goals, so that humans and AI empower each other through mutually beneficial goals.
We define empowerment as the degree of autonomy and self-determination of people and communities. AI should help people become more self-sufficient, more confident in controlling their lives, and more able to influence the world around them.
Our approach
We take a human-centered approach to the design of AI: people are involved in every step of the workflow, from defining goals and producing data to evaluating outcomes. Our work is use-inspired: projects start from concrete settings where AI can provide value, from crowd work platforms and research laboratories to search-and-rescue teams and classrooms. And it aims at social good: technology designed to rebalance power, reduce friction and promote well-being.
We develop models and methods that let machines learn from humans while accounting for cognitive biases and preferences, and we optimize AI systems according to human feedback. We perform real-world experiments to test the feasibility of current technologies in the wild, explore their limitations, and define mitigation strategies to create accountable solutions.
The lab's framing is developed in the director's book Human-AI Empowerment: An Interdisciplinary Perspective (CRC Press, 2025), which proposes methods for studying and optimizing AI's long-term impact on human empowerment, and extended to automation in Human-Centered Automation (CRC Press, 2026).
Milestones
Our story so far
The Human-AI Empowerment Lab is founded in the Human-Centered Computing Division of Clemson University's School of Computing.
The lab is awarded NSF AIMing: Interactive Conjecture Proving, led by Dr. Toxtli as PI.
Cecilia Delgado Solorzano's A Study of LLM-Powered Student Query Support, developed with Clemson University Libraries, receives the Best Paper Award at ACM MexIHC 2024.
Book Human-AI Empowerment: An Interdisciplinary Perspective, the framework behind the lab's name, is published by Chapman & Hall/CRC.
The lab presents three extended abstracts at ACM CHI 2026 in Barcelona, where Dr. Toxtli serves as Associate Chair.
The director's second book, Human-Centered Automation, is published by CRC Press / Taylor & Francis.
Wangfan Li's Supervisory Control Theory for LLM Revision is published at the inaugural ACM Conference on AI and Agentic Systems (CAIS 2026).
Three papers from the NSF AIMing project will appear at MathNLP 2026, co-located with EMNLP 2026 in Budapest: the NaturalPRISM premise-retrieval benchmark, a context ablation for conjecture generation, and robust procedural mathematical reasoning.
Nine papers accepted at seven NeurIPS 2026 workshops, on auditing agent and AI evaluations, peer review in the age of AI, supervising teams of LLM agents, social simulation and mathematical reasoning, including Fateme Mazdarani's study of mathematical creativity in formal proof generation (MATH-AI).
Visit & contact
School of Computing, Human-Centered Computing DivisionMcAdams Hall, Clemson University
Clemson, SC 29634, USA