Social Choice for AI Ethics and Safety 2027 will take place on February 22 or 23 (TBD), 2027 in Montréal, Canada at AAAI-27.

The workshop is organized by Niclas Boehmer, Vincent Conitzer, Jobst Heitzig, and Wesley Holliday

Submission deadline: November 20, 2026 (AoE). See below for submission requirements.

Notification: December 2, 2026

Submission Site: https://openreview.net/group?id=AAAI.org/2027/Workshop/SC4AI

Description of Workshop

AI ethics and safety involve aligning AI systems with the values and preferences of their owners, users, and people affected by their actions. Yet human values and preferences are themselves diverse and often conflicting. Many approaches to AI alignment therefore involve explicitly or implicitly aggregating information about different people’s preferences and values. Social choice theory provides formal tools for preference aggregation, along with rigorous methods for analyzing the properties of these mechanisms.

SC4AI’27 brings together researchers from social choice, AI alignment, AI ethics, and related fields to explore how social choice concepts and methods can contribute to the development of safer, more ethical, and more pluralistic AI systems.

Topics

Topics of interest include but are not limited to: social choice methods relevant to AI alignment and collective human feedback; pluralistic alignment; collective approaches to AI steering and oversight; formal representation and aggregation of values, ethical principles, and incomplete preferences; mechanism design for human–AI and AI–AI interaction; scalable preference elicitation and collective-choice tools; and related economic-theoretic approaches to AI ethics and safety.

Format of Workshop

SC4AI’27 will be a one-day, non-archival workshop combining contributed talks, interactive and breakout-group discussions, and a concluding panel. A poster session may be added depending on the number of submissions.

Attendance

Attendance is open to all interested participants. We particularly welcome participants from the social-choice and AI ethics and safety communities.

Submission Requirements

We invite full papers (6–8 pages) and short papers (2–4 pages), excluding references and appendices, using the AAAI format. Submissions may present completed work, work in progress, position papers, or promising research directions relevant to the workshop. Authors may choose whether to anonymize their submissions. The workshop is non-archival, and submissions of previously published or concurrently submitted work are welcome.