Photo: Marc-Olivier Jodoin / Unsplash
HPAI@AAAI2027 Palais des congrès, Montreal, Canada Feb 16–17, 2027 (Exact date TBD)

Homogenization in AI

Social Impacts, Pluralism, and Agents

Bridge Program AAAI 2027

AI systems are advancing towards greater autonomy, interacting directly with users and even one another. As these systems become embedded in the real world, they are not merely executing individual tasks but also forming interconnected ecosystems. These ecosystems increasingly shape whose perspectives and interests are represented, which decisions are made, and who benefits from or bears the consequences of those decisions.

This bridge program addresses critical gaps at the intersection of homogenization and pluralism in AI, with an emphasis on its social impacts in the rapidly evolving landscape of agentic systems. We will bring together researchers from various disciplines to investigate how homogenization emerges across interconnected ecosystems, understand its social impacts, and explore how pluralism can help preserve diverse perspectives, values, and interests.

Disciplines & topics

AISocial sciencePhilosophyAlignmentResponsible AIHCI
Call for Papers View the schedule

Challenges

The bridge is organized around three related challenges: understanding the social impacts of homogenization, promoting pluralism, and studying how these concerns develop in agentic systems.

I

Challenge I: Social Impacts of Homogenization

  • What social harms emerge when AI systems become increasingly homogenized? Which communities are most vulnerable to homogenization in AI?
  • How can we operationalize metrics for homogenization of creative expression, language preferences, downstream utility, etc., going beyond conventional discrete predictive settings?
II

Challenge II: Promoting Pluralism in AI

  • How can diversity be preserved and promoted in the face of increasingly homogenized AI systems, while maintaining the benefits that scale may provide?
  • To what extent do shared pre-training datasets, architectural design choices, and post-training or fine-tuning on overlapping benchmarks contribute to downstream homogenization in AI systems, and how can we prioritize pluralism under such shared infrastructures?
III

Challenge III: Homogenization in Agentic Systems

  • How can agent interactions further amplify shared failures, create conformity towards the majority, and polarize otherwise moderate positions? What harms and social impacts emerge from such failures?
  • In what ways can pluralism in AI systems be encouraged and preserved under the pressures of conformity and sycophancy in agent interactions?

Call for Papers

Topics of Interest

We welcome submissions on topics including, but not limited to:

  • Social impacts of AI homogenization
  • Metrics and evaluation for homogenization
  • Cultural and linguistic homogenization
  • Homogenization in creative expression
  • Multiplicity and pluralism in AI
  • Preserving diversity in AI infrastructures
  • Data, architectures, and training practices as drivers of homogenization
  • Collective failures and emergent behavior in multi-agent systems
  • Methods for fostering pluralistic and diverse AI ecosystems

Research Track

6–9 pages

Novel contributions in full length papers. All accepted papers will be presented as posters, while selected papers will also be invited for 5-minute contributed talks.

Work already published in an archival venue will not be accepted.

Proposals and Perspectives Track

up to 3 pages

Interdisciplinary proposals and community perspectives that critically examine current approaches or highlight alternative priorities for the community. Selected proposals will be given 3-minute proposal talks to help researchers attract collaborators.

Fast Review Track

no page limit

We invite papers previously submitted to other venues and reviewed, whether accepted or rejected. If you believe it is relevant to the workshop, submit the paper with its reviews; selected papers will be invited for presentation as posters.

Important dates All deadlines are 11:59 PM Anywhere on Earth (AoE)

Submission Deadline (for all tracks)
Nov 22, 2026
Acceptance Notification
Dec 14, 2026

Submissions will open soon.

Submission Format

Submissions use the AAAI two-column format. See the AAAI-27 Author Kit for style files and the AAAI-27 submission instructions for formatting guidelines. Page limits exclude references and appendices.

Dual Submission Policy

We invite submissions of ongoing unpublished work as well as work submitted elsewhere currently under review. However, if the paper is accepted at a parallel venue while undergoing review process at our workshop, it will be moved to the fast review track, including papers accepted at the AAAI main track. Please inform us if a decision has been received for your paper in a parallel venue during the review timeline.

Generative AI Policy

For the Research Track, we will follow AAAI's policy on Appropriate use of Generative AI by Authors.

For the Proposals and Perspectives Track, we will apply stricter screening to AI-generated or AI-assisted submissions. The track aims to provide a platform for researchers in the community to share their perspectives. We want to hear from you, not an AI generated submission.

Schedule

Coming soon

Organizers