1. Context

Artificial Intelligence has rapidly become a pervasive component of formal and informal education, with students, teachers and science communicators increasingly relying on generative tools to produce text, images, diagrams and learning materials—often without adequate guidance or awareness of associated risks. For science and specifically for astronomy education and public engagement, this dual reality presents significant opportunities and serious challenges.

AI systems such as Large Language Models (LLMs), multimodal image generators, speech-to-text and text-to-speech technologies, adaptive learning platforms and evidence-based analytics can greatly enhance access, inclusion and personalization. At the same time, AI-generated content and visuals can propagate inaccuracies, biases and misinformation, potentially undermining scientific understanding and trust.

In astronomy, these risks are amplified by the central role of imagery in teaching and public engagement. AI-generated astronomical visuals, while often aesthetically compelling, may unintentionally reinforce misconceptions about scale, physics and observational limitations (for example introducing impossible physical structures or geometries, incorrect or unclear colour mappings or astronomical objects that violate basic astrophysical constraints).

Currently, there is also a scarcity of peer-reviewed studies examining how students interpret AI-generated explanations or visualisations and assessing their effectiveness and risks. In this context, communicators, educators and scientists need to exercise additional caution and skepticism when working with AI-generated content to avoid unintentionally propagating incorrect information.

This tension is amplified by fragmented international policies: the EU’s AI Act enforces strict transparency rules; the United States operates within a heterogeneous, decentralised regulatory environment; China applies strong national controls; Japan integrates AI formally in school curricula; while many countries lack guidance altogether. For a global community like the IAU, these discrepancies highlight the urgent need for shared understanding and coordinated action.

2. Objectives and functions of the Meeting

The proposed Focus Meeting addresses these challenges by convening astronomers, educators, outreach professionals, AI researchers, ethicists and policymakers to:

• map current uses of AI in astronomy education and outreach;
• evaluate opportunities for learning, pedagogy, accessibility and innovation;
• identify risks related to scientific accuracy, bias, ethics and policy;
• compare regulatory frameworks and their implications for educational practice;
• showcase case studies and tools from diverse regions;
• collaboratively develop preliminary Recommendations and good practices from the International Community.

To fulfil these goals in a rapidly evolving field, the programme integrates three scientific functions: Analyze the current state of AI in Astronomy education and outreach through invited overviews on opportunities, challenges and policy differences, also contributing to mapping research gaps and identifying priority areas.

Gather community insights on what works, what does not work—and why—through dedicated sessions showcasing tools, projects and applications developed in diverse cultural and socioeconomic contexts.

Structured discussion and collective synthesis through a final panel to identify converging themes, open questions and gaps, leading to the preparation of a Community Recommendations document This structure reflects the interdisciplinary nature of the topic and ensures effective dialogue among educators, communicators, AI experts, museum and planetarium professionals, and policy specialists. A hybrid format and proactive SOC engagement with early-career researchers further ensure global inclusiveness and accessibility.