About the workshop

Scope and Aims

Over the last twenty years, the use of AI in low- and high-stakes decision-making systems has increased dramatically.  This widespread adoption has led to present-day concerns around AI misalignment. Many turn to human alignment of AI systems to address these concerns, aiming to reduce undesirable or harmful behaviors in AI and make them more aligned with human values and characteristics. Although there is no universally accepted definition of, or approach to, alignment, it is broadly accepted as the goal and process of making AIs behave in line with human intentions and values.

The goal of this workshop, organized in support of DARPA’s In the Moment (ITM) research program, is to present engaging and high-impact challenges to the community by formulating and discussing key fundamental questions from the perspectives of computer science, artificial intelligence, psychology, and the broader social sciences. Our workshop will address several important questions that have been raised in the research community, for example, is it possible to align AI to human values? If so, which values and/or attributes should we be aligning to? How does alignment work across and between individuals and levels of an organization?  How can we align AI to humans in ways that are more succinct and more reliable than currently found in AI literature?  How might human alignment increase human’s likelihood to trust and/or delegate responsibility to AI? Why or why not? If so, how do we get there?  What are the ethical, legal, and societal implications of human alignment? 

Content and Objectives

The workshop will solicit contributions in the form of paper submissions and invited talks covering whether, and if so, how, human aligned AI will lead to more trustworthy and better performing AI systems, including, but not limited to:

1. Methods and Theory

   1. Theory of alignment

   2. Arguments/positions for and against human aligned AI 

   3. Models that characterize and quantify human decision-making and values for decision-making performance and trust in AI

   4. Models designed from trusted decision-makers

   5. Analysis of human decision-making processes and frameworks and relation to trust in other people/AI and performance

   6. Metrics for associating alignment with trust

   7. Metrics or quantification of AI alignment

2. Application

   1. Applied research use cases

   2. Ethics, risk, and fairness of human aligned AI decision-makers

   3. Laws, regulation, and policy for human aligned AI decision-makers

The call for papers can be found here.

Reach out to davidchan@berkeley.edu or ewartdevisser@gmail.com if you have any questions about the workshop.

2025 IEEE Conference on Artificial Intelligence (IEEE CAI) Website: https://cai.ieee.org/2025/

Organizing Committee

Program Committee

Neil Shortland, Ph.D. - UMass Lowell

Amy Summerville - Kairos Research

Jennifer McVay - CACI, Inc

Ritwik Gupta - UC Berkeley 

Brian Hu - Kitware

Christopher Funk - Kitware

Spencer Kohn - Perceptronics Solutions, Inc. 

Colin Holbrook - University of California, Merced

Alice Leung - BBN

Rosina Weber - Drexel University

Joseph Borders - Shadowbox Training

Anik Sen - Drexel University

Jacob Hyde - Warfigher Consulting