The 2nd Workshop on AI for Urban Planning aims to bring together researchers, practitioners, and policymakers to explore innovative AI-driven solutions for the multifaceted challenges in urban planning. The rise of new technologies and diverse data sources has produced vast, multimodal urban datasets, raising challenges related to data quality, ethical considerations, and the interpretability of AI models. This workshop will explore how AI and urban science can foster a co-learning paradigm—the "New Urban Science"—to build smarter, more equitable, and sustainable cities.

The workshop is held in conjunction with the 40th Annual AAAI Conference on Artificial Intelligence (AAAI-2026), one of the world's leading AI conferences. Organized by the Association for the Advancement of Artificial Intelligence, AAAI is a key platform where researchers, practitioners, and industry experts present groundbreaking advancements and explore emerging trends in AI. This partnership provides the workshop with a vital forum for interdisciplinary knowledge exchange, fostering collaboration between AI researchers, urban planners, and policymakers.

This year, we welcome submissions across three tracks. In addition to the Regular Track for research papers covering broad AI and urban planning topics, we have introduced two specialized tracks to foster community-wide collaboration: Task and Problem Definition aimed at establishing shared standards and benchmarks, and Tutorial presentations showcasing practical tools and methodologies.

Track 1: Regular Paper

We encourage submissions on a broad range of topics related to AI in urban planning, including but not limited to:

Representation and Quantification of Urban Environments:

• Representation learning for spatio-temporal data
• Multimodal data fusion
• Graph neural networks in representing urban forms
• Multi-view learning
• Domain shift & generalization

Predictive Modeling with Urban Data:

• Urban time series forecasting
• Spatio-temporal forecasting (e.g., energy, traffic, crowd flow)
• Demographic prediction
• City climate modeling

Generative Modeling, AIGC, and Large Models for Urban Planning:

• Urban form design
• Land use configuration
• Architecture design
• Landscape design
• Transportation system design (e.g., road network)

Evaluation and Simulation:

• Reinforcement learning powered simulation
• LLM agent for evaluation and simulation (i.e., stakeholder role-play)
• Human-in-the-loop simulation
• LLM-in-the-loop simulation

Ethics of AI in Urban Planning:

• Fairness and bias
• Community engagement
• Transparency and accountability
• Data privacy
• Interpretability

Applications, Systems, and Tools:

• Urban digital twins
• AR, VR, and metaverse for urban planning
• Carbon neutrality
• Urban utility/resources allocation
• Planning support systems for policy makers

Track 2: Task & Problem Definition

Currently, the field of AI for urban planning lacks unified problem settings, standardized tasks, and common datasets, which can make it difficult to compare methods and measure progress. This track invites constructive suggestions and visionary proposals aimed at building a strong, shared benchmark for the community.

We welcome contributions that:

• Propose new standardized tasks for AI in urban planning
• Define comprehensive evaluation frameworks
• Present ideas for creating shared datasets to accelerate innovation
• Suggest methodologies for fair comparison of different approaches

Track 3: Tutorials

This track invites researchers and practitioners to introduce their libraries, tools, and systematic works to the workshop attendees. The goal is to provide a platform for hands-on demonstrations and in-depth presentations of software and systems that are valuable to the AI and urban planning community.

We encourage proposals that:

• Demonstrate practical software tools and libraries
• Provide hands-on tutorials for AI urban planning applications
• Share systematic methodological frameworks
• Help disseminate practical knowledge and useful tools to the community