Call for Papers: PANDORA Workshop @ EMNLP 2026
Workshop Title: PANDORA: Pluralistic AI & NLP: Diversity-aware, Sociotechnical, Responsible Alignment
Workshop Date: TBD (with EMNLP 2026, October 24–29, Budapest, Hungary)
Format: Hybrid (Online + In-Person)
Workshop Description
The PANDORA workshop focuses on pluralism in AI and NLP: how to design, train, evaluate, and govern language technologies in ways that account for diverse human judgments, values, and social contexts. As AI systems are increasingly trained and evaluated using human feedback, it has become clear that disagreement among people is often not just annotation noise, but a meaningful signal reflecting differences in social position, culture, language background, and lived experience.
Traditional NLP pipelines often optimize for a single “gold” target, collapsing disagreement into consensus. But in many deployed tasks—such as toxicity detection, politeness, stance, and summarization quality—variation in judgments is systematic and legitimate. Ignoring such pluralism is not only a technical limitation, but also a sociotechnical failure: it obscures whose values are represented in models and evaluation protocols, especially in alignment pipelines based on aggregated human feedback.
PANDORA aims to advance this discussion end-to-end, spanning:
- Data: eliciting and representing pluralistic judgments, and distinguishing error from meaningful variation
- Modeling & Alignment: learning from diverse and conflicting feedback without collapsing to a single “best” view
- Sociotechnical Evaluation: assessing systems in pluralism- and context-aware ways beyond single-ground-truth metrics
- Normative Questions: understanding when pluralism is desirable, when consensus is required, and what trade-offs arise in practice
The workshop also aims to support the development of community standards, shared practices, and actionable guidelines for pluralism-aware NLP and AI, including a potential Pluralism-aware Reporting Checklist for datasets, preference data, and evaluations.
The workshop is intended as a venue for researchers and practitioners working across NLP, machine learning, AI alignment, HCI, computational social science, philosophy, and policy.
Submission Guidelines
We invite submissions in the following categories:
- Long Papers: Up to 8 pages of content, plus unlimited references and appendices
- Short Papers: Up to 4 pages of content, plus unlimited references and appendices
- Extended Abstracts: Up to 2 pages of content, plus unlimited references and appendices
Authors may choose whether their submission is archival or non-archival.Non-archival submissions may include previously published work, work in progress, or emerging ideas. We also welcome interdisciplinary, conceptual, and community-grounded work, including submissions that focus on a single language, dialect, or social setting. In addition to standard paper categories, PANDORA welcomes Reflections & Provocations: short, well-argued pieces offering a critique, viewpoint, or provocative perspective on pluralism in NLP and AI.
Submission Requirements
Submissions must adhere to the official ACL style templates. Papers must be in two-column ACL format and submitted as PDF files.Submissions should be anonymized for review. Submissions will be handled via OpenReview. Please note that submissions must be fully self-contained. Supplementary material may be included only in the appendix, and reviewers are not required to review it.
Topics
Submissions may address a broad range of topics related to pluralism-aware NLP and responsible alignment, including but not limited to:
- Human label variation and perspectivist NLP
- Non-aggregated annotation and disagreement-aware data collection
- Aggregation and representation of annotator disagreement
- Annotator-aware, group-aware, and uncertainty-aware models
- Pluralistic AI alignment and diverse human values
- Multi-objective and preference-conditioned alignment
- Disagreement-aware post-training and learning objectives
- Pluralism-aware evaluation frameworks and metrics
- Calibration, uncertainty, fairness, and inclusivity in evaluation
- Decision policies under pluralism: abstention, deferral, personalization, and multiple outputs
- Participatory design and human-centered approaches to pluralistic AI
- Governance, policy, and normative analysis of alignment under disagreement
- Applications and case studies in moderation, education, health, multilingual, cross-cultural, and multimodal systems
Multiple Submission Policy
For archival direct submissions, papers must not be under review at another journal, conference, or workshop during the workshop review period, and must not overlap substantially with work that is already published or under review elsewhere.
For non-archival submissions, authors may submit work that is previously published, under review elsewhere, or intended for future submission to another venue, provided that the submission is clearly marked as non-archival.
ARR Commitment
In addition to direct submissions, PANDORA expects to welcome ACL Rolling Review (ARR) commitments, subject to the workshop and host conference timeline. Papers that have already received ARR reviews and a meta-review may be considered based on those existing reviews together with fit to the workshop scope.
Submission Links
- Direct Submission: TBD
- ARR Commitment: TBD
Important Dates
All deadlines are 11:59 PM UTC-12:00 (“Anywhere on Earth”)
- Direct Submission Deadline: July 31, 2026
- ARR Commitment Deadline: August 28, 2026
- Acceptance Notifications: September 11, 2026
- Camera-ready Deadline: September 25, 2026
- Workshop Date: During EMNLP 2026, October 24–29, 2026
Contact Information
- Workshop Website: http://pandora-workshop.github.io
- X: https://x.com/Pandora_EMNLP26
- Bsky: https://bsky.app/profile/pandora-workshop.bsky.social
- Email: pandora.workshop.1@gmail.com
——————-——————-——————-——————-——————-——————-————-
Key Dates from EMNLP 2026
ARR submission - May 25
Commitment - Aug 2
Decision - Aug 20
Camera ready - Sep 20
Conference - Oct 24 - 29
PANDORA Dates
Direct Submission - July 31
- Reviewer allocation - August 3
- Deadline - Aug 24
- Reviewer chasing - Sep 2
ARR Commitment - Aug 28
Acceptance Notifications - September 11
Camera Ready Deadline: September 25