Call for papers GenAI4SocialGood Workshop@CIKM'26

[Marco Minici]

================================================================================
                    CALL FOR PAPERS — GenAI4SG @ CIKM 2026
       1st Workshop on Generative AI for Social Good and Vulnerable
                              Populations
================================================================================


OVERVIEW
--------------------------------------------------------------------------------
Generative AI now functions as a general-purpose engine for producing text,
images, code, and grounded explanations across knowledge-intensive tasks.
That capability lands hardest where it has been needed longest — in domains
where expertise, data, and resources have always run thin, from rare-disease
diagnosis and low-resource education to accessibility, crisis response, and
public-service access. In these settings, generative systems can extend
expert capacity into rooms it never reached before, turning scarce specialist
knowledge into something a frontline worker, a teacher, or a patient can
actually use.

The same conditions that make GenAI valuable for social good also make its
failures dangerous. Hallucination, linguistic and cultural bias, privacy
leakage, and computational costs that price out under-resourced organizations
compound quickly in these high-stakes settings, and the populations meant to
benefit often end up carrying the risk. Responsible design is what turns this
technology's promise into something deployable — the two have to be built
together.

This workshop brings together interdisciplinary researchers and practitioners
to interrogate both sides of that equation: how retrieval-augmented, agentic,
multimodal, and knowledge-grounded generative AI systems can be built,
evaluated, and deployed in service of underserved and vulnerable communities,
in line with the UN Sustainable Development Goals. The workshop draws on the
CIKM community's core strengths — information retrieval, knowledge
management, data mining, machine learning, NLP, and recommender systems — to
retrieve evidence, build and maintain knowledge bases, generate grounded
explanations, evaluate information quality, and deploy systems over
heterogeneous data.


TOPICS OF INTEREST
--------------------------------------------------------------------------------
Topics of interest include, but are not limited to:

  • Retrieval-augmented generation and evidence-grounded GenAI for
    high-impact domains
  • Knowledge-graph integration, semantic grounding, fact-checking, and
    source verification
  • LLM agents and agentic workflows for social-good applications
  • Multimodal GenAI for health, education, accessibility, crisis response,
    and sustainability
  • Synthetic data generation, data-centric GenAI, and low-resource
    adaptation
  • Benchmarks, datasets, and evaluation protocols for societal benefit and
    deployment risk
  • Fairness, robustness, privacy, explainability, and human-centered
    assessment
  • Efficient, sustainable, and practical deployment for public-interest
    organizations
  • Applications supporting vulnerable and underserved populations
  • Information integrity, misinformation detection, and coordinated
    manipulation analysis

  ★ Deployment & Impact Focus:
    Submissions are especially encouraged to demonstrate real-world
    deployability — including low-resource and efficient methods,
    human-centered evaluation, and explicit attention to risks and benefits
    for vulnerable or underserved populations — alongside technical
    effectiveness.


INSTRUCTIONS FOR SUBMISSION
--------------------------------------------------------------------------------
The workshop welcomes:

  1. Full papers presenting mature technical or empirical contributions
  2. Short and position papers presenting preliminary results or research
     agendas
  3. Dataset, benchmark, and demo papers presenting reusable resources or
     systems

Submitted papers will undergo peer review coordinated by the workshop's
Program Committee, with each submission receiving at least two reviews.

Evaluation criteria include:
  - Relevance to the workshop
  - Technical quality
  - Novelty
  - Clarity
  - Methodological soundness
  - Potential societal benefit
  - Awareness of risks for vulnerable or underserved populations

Format & logistics:
  - Papers should be prepared in the ACM two-column format.
  - Submissions will be handled electronically through EasyChair or a
    similar system (submission link to be announced).
  - Workshop papers will not be included in the ACM proceedings by default;
    archival publication options will be announced in accordance with CIKM
    policies.
  - At least one author of each accepted paper is expected to register for
    CIKM 2026 and present the work on-site in Rome.


KEY DATES  (11:59 PM AoE unless noted)
--------------------------------------------------------------------------------
  Tentative paper submission deadline .... September 13, 2026
  Tentative acceptance notification ...... September 23, 2026
  Camera-ready papers .................... TBD
  Workshop day ............................ November 8, 2026
                                            (held with CIKM 2026, Rome, Italy)

================================================================================