This panel explores the evolving role of self-adaptive systems in the context of the AI era, addressing critical themes such as defining system boundaries, adaptation decision-making (including the role of Large Language Models), runtime uncertainty handling, and the integration of human roles and ethical considerations (Fairness, Accountability, Transparency). The panel seeks to discuss how these aspects influence research directions and the broader impact of self-adaptive systems, as well as how the SEAMS community wishes to adapt its research agenda in response to recent advances in AI technologies.

The panel will be conducted in a dynamic “fishbowl” format:

• A group of invited panelists initiates the discussion.
• At least one open seat is available from the start, allowing audience members to join the conversation. Audience participants can introduce new questions, provide responses, and engage directly with panelists.
• Audience members occupying an open seat are encouraged to step away once they’ve contributed their insights, enabling continuous rotation.
• Invited panelists may also leave their seats when they feel they have concluded their contributions, further opening seats to audience participation.