Human-Centric AI:

This workshop aims to bring together researchers and practitioners from foundational areas of Human-Centric AI—such as Explainability, Trustworthiness, Fairness, and Privacy — to build a shared understanding of its core principles and challenges.

It is imperative that AI systems operate in ways that are transparent, trustworthy, and ethically sound. Our vision of Human-Centric AI is built on four foundational pillars: explainability, trustworthiness, fairness, and privacy. Developing truly human-centric AI requires more than technical innovation—it demands interdisciplinary collaboration. Diverse perspectives are essential for tackling the complex and nuanced challenges in the field. To this end, fostering dialogue among researchers and practitioners across disciplines is vital for advancing shared values at the core of human-centric AI.

This workshop will explore key challenges and emerging solutions in the development of human-centric AI, with a focus on explainability, trustworthiness, fairness, and privacy. We welcome not only theoretical advances but also applied research and case studies that show how human-centered principles are realized in real-world AI applications.

Target Audience

Our intended audience includes researchers, graduate students, practitioners, and policy makers from around the world who are interested in Human-Centric AI, particularly in areas such as Explainable AI, Trustworthiness, Fairness, and Privacy.

Topics of interest include, but are not limited to:

• Explainable AI (XAI) methods and evaluations
• Trustworthy AI, including Robustness and Safety
• Fairness-aware machine learning
• Privacy-preserving machine learning
• Model interpretability (for Large Language Models and Large Multi-modal Models)
• AI Ethics, Frameworks, and Governance
• Global AI regulations (e.g., EU AI Act, Korea's AI Basic Act) and their practical implications
• Case studies of human-centric AI in real-world applications
• Human-AI Interaction

Important Dates

All deadlines are at 11:59PM UTC-12:00 ("Anywhere on Earth", AoE)

* Note: You will require an EasyChair account to submit.

Paper Submission

Manuscripts should be submitted via the CIKM 2025 EasyChair site (https://easychair.org/my/conference?conf=cikm25) in PDF format, using the 2-column ACM sigconf template. Please refer to the official ACM proceedings template available at https://www.acm.org/publications/proceedings-template .

We accept both short and long paper submissions. Short papers are limited to 4 pages, and long papers to 9 pages, including any appendix. Authors may include an appendix only if it fits within the page limit; there is no additional space allowed beyond the limit for appendices. References may extend to an unlimited number of pages and do not count toward the page limit.

Papers will be evaluated on originality, impact, relevance, and clarity through a single-blind review process conducted by 2 to 3 program committee members. Based on review scores and topical balance, accepted papers will be assigned to either oral or poster presentations.

Submissions are non-archival. At least one author of each accepted paper must register for CIKM 2025 and present the paper on-site in Seoul, Korea, as scheduled in the program.

Organizers

Jaesik Choi

KAIST, Korea

Bohyung Han

Seoul National University, Korea

Myoung-Wan Koo

Sogang University, Korea

Kyoungman Bae

ETRI, Korea

Chang D. Yoo

KAIST, Korea

Simon S. Woo

Sungkyunkwan University, Korea

Wojciech Samek

TU Berlin and HHI, Germany

• Jaesik Choi, KAIST, Korea
• Bohyung Han, Seoul National University, Korea
• Myoung-Wan Koo, Sogang University, Korea
• Kyoungman Bae, ETRI, Korea
• Chang D. Yoo, KAIST, Korea
• Simon S. Woo, Sungkyunkwan University, Korea
• Nari Kim, KAIST, Korea
• Maximilian Dreyer, Fraunhofer HHI, Germany
• Hwaran Lee, Sogang University, Korea
• Sangwoo Heo, NAVER AI Risk Management Center, Korea