Generative AI for Hypothetical Exoplanet Ecosystems: A Physics-Informed Generative Framework for Space Discovery

We propose a physics-informed generative framework to simulate plausible exoplanet biospheres conditioned on stellar type, orbital/planetary parameters, and atmospheric chemistry. The model synthesizes ecosystem state variables (energy flow, trophic structure, metabolism classes) under thermodynamic and radiative-transfer constraints. Results suggest that generative AI can act as a hypothesis engine to guide biosignature searches.