Machine learning, statistical learning, and artificial intelligence (hereafter, all considered “AI”) are actively transforming how ecologists work with data. However, there are also rich and growing opportunities for AI to inform theoretical ecology. The applications of AI in theoretical ecology are diverse. For example, AI can help us better understand governing ecological mechanisms such as represented in ecological models. In this application, AI represents a shift from deductive to inductive or hybrid approaches to developing scientific theories. Ecological models have long been generated deductively, by hypothesizing mechanisms, generating theoretical models that embody those mechanisms, and comparing the models against data. However, AI approaches allow for an inductive approach where ecological theory (e.g. ecological models or other theoretical representations) is generated directly from data. This has the potential to confer more insights into governing ecological mechanisms. Other approaches use Markov decision processes to optimize decision-making regarding ecological systems under challenging uncertainties. The purpose of this symposium will be to explore the common challenges and opportunities for research that uses AI to inform theoretical ecology in these endeavours. This session will showcase (i) AI approaches for modeling species distributions to enable the construction of niche models (ii) how AI can help us understand the impact of relaxing common simplifying assumptions of classical ecological theory, (iii) using AI to inform decision-making for conservation science under imperfect knowledge and resource constraints, and (iv) AI tools for gaining insight into complex nonlinear ecological dynamics. The symposium audience will learn from these examples how AI is improving both our theoretical understanding as well as the precision and accuracy of theoretical models of ecological systems and the quality of conservation decision-making. In the symposium we will hear from four researchers involved in this rapidly growing subfield of ecology, followed by an opportunity for the panel to interface with the audience through Q&A and discussion. The connection of our symposium to the theme of “Harnessing the ecological data revolution” comes from the use of large datasets in many AI approaches, including some of the approaches to be presented in the symposium.