CI 2026  /  Submit  /  Accepted Workshops

Accepted Workshops

Important Dates

List of Accepted Workshops

Click on any workshop title below to view its full details.

Full-day Workshops

Half-day Workshops

Workshop Details

Human-AI Teaming: Useful Framework or Empty Metaphor?

Format: Full day

Website: https://hcomp-ci2026-hat.github.io/

Organizers:

  • Sam Moradzadeh (Penn State University, USA)
  • Andrew M. Sherrill (Emory University, USA)
  • Rosa I. Arriaga (Georgia Institute of Technology, USA)
  • Christopher W. Wiese (Georgia Institute of Technology, USA)
  • Jung Ah (Julie) Lee (Georgia Institute of Technology, USA)
  • Sadie K. Hogge (Georgia Institute of Technology, USA)
  • Saeed Abdullah (Penn State University, USA)

Abstract: "Human-AI teaming" has become one of the most widely invoked concepts in AI research, design, and policy. Yet beneath the surface of this appealing phrase lies a set of unresolved questions that the research community has largely deferred. Rather than asking whether teaming is the correct term, this workshop asks a more generative question: what work does the teaming construct actually do, and for whom? For researchers, it may offer a frame for studying coordination and division of labor. For designers, it may suggest interaction structures that go beyond tool use. For end-users, it may set expectations, helpful or misleading, about agency, accountability, and trust. We bring together researchers from human-computer interaction, cognitive science, clinical studies, and AI to examine the functions and limits of the teaming concept, compare it against alternative framings such as augmentation, delegation, and scaffolding, and work toward a more precise vocabulary for describing and designing expert-AI collaboration. Through short provocations, structured debates, and collaborative framework-building, the workshop will produce a shared position paper capturing points of agreement and productive disagreement, with the goal of grounding future CI and HCOMP research in a more honest and rigorous account of what human-AI collaboration actually is, and what it should aspire to be.

The First Workshop on Sovereign AI for Collaborative and Pluralistic AI Ecosystems

Format: Full day

Website: https://sovereignai-workshop.github.io/

Organizers:

  • Lun-Wei Ku (Academia Sinica, Taiwan)
  • Wan-Jhen (Crystal) Wu (Academia Sinica, Taiwan)

Abstract: Sovereign AI is not only about where models are developed, deployed, or hosted, nor only about who owns or controls them. It is about who has the power to shape, question, and govern AI systems when they act on behalf of individuals, communities, and institutions. AI systems are increasingly embedded in public services, media, civic participation, and everyday decision-making. Yet many of these systems are designed and operated by a concentrated set of technical and commercial actors, leaving affected people with limited visibility into how they work and little ability to contest or change them. Genuine sovereignty requires more than participation or consultation: it requires that people, communities, and institutions be able to examine AI systems, challenge their behavior, and, when they fail to meet local needs, values, or legal standards, require that they be corrected, reconfigured, or replaced. What matters is not only the ability to be heard, but meaningful authority over outcomes. This is both a technical and a governance challenge. It asks how humans and AI can complement one another without reducing people to passive users or treating automation as inevitable. Addressing this challenge requires connecting AI/ML, HCI and design, institutional governance, law and policy, and journalism, while moving affected communities into active roles in rule-making, oversight, and evaluation. This workshop brings together researchers and practitioners across these fields to examine real-world deployments, develop cross-disciplinary Sovereign AI Problem Maps, and produce actionable recommendations. Through keynotes, paper presentations, and posters, it works in the spirit of the HCOMP × CI 2026 theme, "Connections." We welcome papers, work-in-progress, position papers, surveys, demos, and deployment case studies.

Past Meets Future: 3rd Annual Workshop on Human-AI Interaction for Digital Humanities and Cultural Heritage

Format: Full day

Website: https://past-meets-future.github.io/2026/

Organizers:

  • Fei Shan (Virginia Tech, USA)
  • David A. Shamma (Northeastern University Oakland, USA)
  • Victoria Van Hyning (University of Maryland, USA)
  • Benjamin Charles Germain Lee (University of Washington, USA)
  • Vikram Mohanty (Carnegie Mellon University, USA)
  • Kurt Luther (Virginia Tech, USA)

Abstract: This workshop aims to bridge the gap between the Human-AI Interaction (HAI) research community and fields like digital humanities, history, and cultural heritage to address critical challenges such as data quality, accessibility, and public engagement. The primary goal is to foster interdisciplinary dialogue to develop human-centered tools and innovative HAI methodologies. Workshop day activities feature participant-led presentations and collaborative discussions focused on building shared understandings and identifying open research opportunities. Building on past momentum, the previous two editions of this workshop series were successfully hosted at ACM IUI 2024 and ASIS&T 2025.

AI CHAOS! 3rd Workshop on the Challenges for Human Oversight of AI Systems

Format: Half day

Website: https://sites.google.com/view/aichaos/hcomp-2026

Organizers:

  • Shreyan Biswas (Delft University of Technology, The Netherlands)
  • Ji-Youn Jung (Carnegie Mellon University, USA)
  • Marie-Therese Sekwenz (Delft University of Technology, The Netherlands)
  • Min Hun Lee (Singapore Management University, Singapore)
  • Harmanpreet Kaur (University of Minnesota, USA)
  • Simo Hosio (University of Oulu, Finland)
  • Ujwal Gadiraju (Delft University of Technology, The Netherlands)

Abstract: As AI systems are increasingly deployed in high-stakes domains, ranging from healthcare and autonomous driving to platform governance and automated legal compliance, their failures directly threaten human safety and fundamental rights. Central to HCOMP’s mission around Human-AI Complementarity and Alignment, human oversight serves as a vital socio-technical safeguard. Drawing on recent interdisciplinary roadmaps, effective oversight is conceptualized not as a solitary interface task, but as a deliberate, risk-mitigating activity structured around the two operational pillars of continuous monitoring and timely intervention. This responsibility is distributed across diverse organizational stakeholders, including frontline operators, domain experts, and compliance officers. However, emerging regulatory mandates like the European AI Act and the Digital Services Act have outpaced our conceptual clarity and empirical methodologies. Poorly designed oversight architectures risk creating a dangerous illusion of control, obscuring accountability without providing humans with genuine causal agency or systemic support. Embracing the HCOMP 2026 theme of “Connections," this workshop establishes vital links across AI, human computation, HCI, psychology, law, and policy to bridge this critical gap.

The Future of Research in the Age of Human–AI Collective Intelligence

Format: Half day

Website: https://sduyr.github.io/hai-ci-workshop-2026

Organizers:

  • Senjuti Dutta (University of Colorado, Boulder, USA)
  • Alex Bentley (University of Tennessee, Knoxville, USA)
  • Chuanren Liu (University of Tennessee, Knoxville, USA)

Abstract: What is the future of research as a human–AI collaboration? Researchers increasingly use LLMs to generate hypotheses, analyze data, synthesize literature, write code, and communicate findings. This raises fundamental questions about how scientific knowledge is produced, evaluated, and shared. Because science functions as a collective intelligence system, understanding these transformations requires both HCOMP perspectives on human–AI complementarity and CI perspectives on distributed cognition and knowledge production. Bringing together researchers from computer science, social science, business analytics, and related fields, this workshop examines changes at multiple scales: from knowledge work and research practice, to scientific communication, to broader questions about the future evolution of collective intelligence in human–AI knowledge systems. By combining perspectives from HCI, computational social science, and evolutionary theory, participants will establish empirical baselines, develop a research agenda, and identify promising directions for future collaboration.

CLARITY: 1st Workshop on Cognitive Labels for Appropriate Reliance and Information Transparency

Format: Half day

Website: https://sites.google.com/view/aiclarity/home

Organizers:

  • Siddharth Mehrotra (Birla Institute of Technology & Science, India)
  • Yoana Ahmetoglu (University College London, United Kingdom)
  • Jessica He (IBM Research, USA)
  • Marios Constantinides (CYENS Centre of Excellence, Cyprus)
  • Abdallah El Ali (CWI & Utrecht University, The Netherlands)
  • Anna Cox (University College London, United Kingdom)
  • Ujwal Gadiraju (Delft University of Technology, The Netherlands)

Abstract: As generative AI systems seamlessly blend reality with fabrication, helping users calibrate their trust becomes a pressing challenge. Without appropriate signals, users may over-rely on incorrect AI advice or hallucinated outputs on one hand, or harbor skepticism toward digital media on the other hand. To foster appropriate reliance and safeguard responsible user opinion formation, emerging mandates like the European AI Act enforce strict requirements for AI disclosures. However, static, compliance-driven labels often induce user fatigue or trigger cognitive effects that make users inadvertently validate unlabeled misinformation. Our workshop aims to establish an interdisciplinary community to address these critical socio-technical gaps. Buoyed by the core theme of HCOMP 2026, this workshop argues to transform AI disclosures from passive regulatory checkboxes into dynamic, interactive tools designed to empower human metacognition to actively foster appropriate trust calibration. We invite full research papers and short position papers exploring the behavioral impacts of disclosure formats, human-centered uncertainty communication, media provenance infrastructure (e.g., watermarking), and strategies to preserve epistemic agency and security in public discourse. Through interactive talks and hands-on co-creation breakout sessions, participants will dissect technical trade-offs and map out human-centric disclosure metrics.

EVOKE: Eliciting Verifiable, Operational Knowledge from Experts

Format: Full day

Website: Coming soon

Organizers:

  • Praveen Paritosh (ML Commons, USA)
  • Vinay K. Chaudhri (Stanford University, USA)
  • Clifton McFate (Cynch AI, USA)
  • Ram Bala (Santa Clara University, USA)

Abstract: We have abundant data and powerful learners. What we lack is a reliable way to elicit knowledge, the structured and reusable claims a person holds, and to prove that what we extracted is useful. EVOKE focuses on knowledge elicitation into representations with formal, provable, or sound inference properties: constraints, optimization models, logic, programs, typed specifications, ontologies, and probabilistic models. It is organized around the one question that separates knowledge from data: How do we know the elicited knowledge is correct and useful? When the target is a formal representation, that question becomes measurable. A solver returns ground truth. A verifier certifies a result. An unsatisfiable core exposes a contradiction the expert did not know they held. The timing is sharp. Large language models have collapsed the cost of the natural-language interface that stalled knowledge acquisition for decades, so we can elicit at scale and in plain language for the first time. But language models elicit into prose, with no guarantee of soundness, and they fail open: a plausible specification that is subtly wrong is the dangerous case. The frontier is eliciting into a form a verifier or solver can underwrite, and measuring whether the captured knowledge holds. EVOKE draws on Leslie Valiant’s educability framework which locates human uniqueness not in learning from examples but in being taught explicit knowledge and chaining it, with belief verification as the safeguard. We convene knowledge engineering, expert judgment, requirements engineering, human-AI interaction, operations research, neuro-symbolic AI, and natural language processing around shared evaluation, a shared task with solver-verifiable ground truth, and the question of when extracted knowledge is provably useful.

Cognitive Security for Human–AI Systems (CogSec)

Format: Full day (hybrid option)

Website: https://noetx.github.io/cogsec-2026/

Organizers:

  • Utsav Gupta (Stanford University, USA)
  • Brett Frischmann (Villanova University, USA)
  • Eric Heng (Stanford University, USA)
  • Andreas Haupt (Stanford Institute for Human-Centered AI; Stanford Digital Economy Lab, USA)

Abstract: As generative and agentic AI become embedded in search, education, companionship, and decision support, a class of trust-and-safety risk is emerging that the field has no shared way to name, test, or govern. A single model output may be accurate and appropriately disclosed. The risk lies instead in how systems shape attention, belief formation, memory, judgment, and agency across many interactions, and in how they reshape collective sensemaking. CogSec proposes cognitive security, the integrity of the conditions under which people and groups reason for themselves, as a research agenda spanning HCOMP (AI-assisted decision-making, over- and under-reliance, oversight) and CI (misinformation, deliberation, civic decision-making). The workshop convenes researchers and practitioners to consolidate a shared vocabulary of cognitive vulnerabilities, advance the measurement of over- and under-reliance, and translate findings into auditable interventions. It is anchored by a hands-on session that builds the Cognitive Vulnerability and Mitigation Catalog (CVMC): a schema pairing each AI affordance and cognitive vulnerability with a candidate mitigation, an evaluation metric, a burden-and-equity check, and a governance hook. Attendees leave having co-authored catalog entries and a community roadmap.

Shaping the Handbook of Human Computation, Second Edition

Format: Half day

Website: https://humancomputation.org/projects/hch2

Organizers:

  • Pietro Michelucci (Human Computation Institute and Cornell University, USA)
  • Caroline Nickerson (Human Computation Institute, USA)
  • Shida Sharif-Bakhtiar (McGill University, Canada)
  • Jérôme Waldispühl (McGill University, Canada)

Abstract: The first edition of the Handbook of Human Computation brought together more than 100 authors from academia, industry, and nonprofit organizations to define and document an emerging interdisciplinary field. Since its publication in 2013, the handbook has been downloaded more than 280,000 times and has served as a foundational reference for researchers, practitioners, and students. Over a decade later, the landscape has changed dramatically. Generative AI, large language models, agentic systems, and increasingly sophisticated human-AI collaboration are transforming how people learn, create, decide, and solve problems together, even as the ACM HCOMP and Collective Intelligence communities converge around designing systems that combine the complementary strengths of humans and machines. This workshop invites the community to help shape the Handbook of Human Computation, Second Edition (HCH2). In general, participants will learn about the handbook vision and editorial process, identify important topics, propose chapters, and contribute to a community-driven roadmap. We will begin wtih a town hall meeting on the evolving role of humans in hybrid intellence systems. Then we’ll ground our thinking with GWAP (Games with a Purpose) case studies that explore the continued value of human contributions in AI-enhanced environments. Then we’ll engage in collaborative ideation with lightning talks as readouts, and conclude with participation pathways for a new publication that will help set a vision for the co-evolution of humanity and AI.

Workshops Co-Chairs

CI 2026 Chair

Vikram Mohanty
Carnegie Mellon University, USA

HCOMP 2026 Chair

Senjuti Dutta
University of Colorado Boulder, USA

For questions about workshop proposals, please contact: hcomp-ci-2026-workshops@acm.org