Submission (via OpenReview) TBA!

The rapid deployment of AI systems in interactive environments marks a shift from isolated prediction to strategic multi-agent interaction. Modern AI agents increasingly operate in settings involving negotiation, competition, and cooperation, yet current machine learning methodologies often lack the theoretical tools to reason about incentives, stability, and long-term behavior in such environments. While game theory provides a principled framework for strategic interaction, its classical formulations struggle to capture the high-dimensional, non-stationary, and data-driven nature of modern learning systems. Conversely, recent advances in AI—particularly large language models—excel at processing rich, unstructured inputs but often lack consistent strategic reasoning.

This workshop aims to bridge this gap by bringing together researchers from machine learning, game theory, and optimization to advance the foundations of learning in strategic settings. We seek to foster dialogue between these communities and identify new tools, abstractions, and benchmarks for understanding and designing multi-agent AI systems.

We invite contributions in the form of: 

• standard and 

• short papers  (new) 

on topics related to game-theoretic learning and strategic AI. Short papers may include preliminary results, benchmarks (or proposals thereof), datasets, desiderata, or position papers. All accepted submissions will be presented as posters, with a subset selected for oral or spotlight presentations.

-> Pages limits: TBA 

**There will be no proceedings; accepted papers will be made available via OpenReview and listed on the workshop website.

Double-blind & Dual submissions

• Submissions must be properly anonymized for double-blind review.

• Papers already accepted at venues with archival proceedings (including the ICML main conference) will not be considered.

• We discourage dual submissions to multiple ICML workshops—please submit to the one that best fits your work.

• Extended abstracts of papers under review at other conferences/journals can be submitted if this is ok for the conference/journal in question (if in doubt, please check with them first). 

Topics:

Topics include, but are not limited to, the following:

• Learning dynamics in strategic settings: How do learning algorithms behave in multi-agent environments? What guarantees can be established beyond equilibrium convergence?

• Algorithmic game theory and game optimization: What new algorithmic frameworks are needed for modern large-scale and non-convex games?

• Mechanism design and adaptive systems: How can mechanisms be designed in settings where learning dynamics shape the equilibrium itself? 

• Stackelberg and asymmetric games: How can we model and control learning in hierarchical or information-asymmetric interactions?

• New solution concepts: Do classical notions (e.g., Nash equilibrium) suffice for modern AI agents, or are new stability concepts needed?

• Strategic reasoning with LLMs and agentic AI: Can language models reliably reason about incentives, beliefs, and interactions in dynamic environments? 

• Multi-agent systems and interaction, incl. Multi-Agent Reinforcement Learning (MARL): How do large populations of learning agents interact, and what risks arise (e.g., correlated failures or monocultures)?

• Benchmarks, datasets, and evaluation: How can we measure strategic capabilities in AI systems beyond static benchmarks?

• Policy, regulation, and societal impact: How can game-theoretic tools inform governance, accountability, and alignment in multi-agent AI systems?

We also welcome negative results, well-motivated ideas that did not succeed, and critical perspectives that highlight missing tools or flawed assumptions, as well as reflections on the impact of agentic AI on the ML community.

Important Dates

• Submission Deadline: April 24, 2026 (AoE)

• Acceptance Notification: TBA

• Camera Ready: TBA

• Workshop: July 2026, Day TBA

• Location: Seoul, South Korea, co-located with ICML

Submission Instructions

• Submission: Will be on OpenReview, link TBA

• Formatting: latex template will be provided -- TBA

• Reviews: Submissions will be evaluated by a reviewing committee. There will be a single round of reviews and no author response.

Questions? Check out the FAQs or reach us at nextgame.icml@gmail.com.