Physics-informed learning was proposed in recent years to combine data-driven techniques (such as neural networks, Gaussian processes, …) with the insights of dynamical systems theory. This hybrid approach benefits from the best of both worlds, with the generalization capabilities of learning methods being underpinned by solid theoretical results. Physics-informed learning has proved to be promising in a wide range of applications, ranging from fluid mechanics to epidemiology, from finance to robotics, and to control. This workshop thus aims at solidifying the role of physics-informed learning for control by collecting different perspectives on the topic. The workshop will consist of five keynote talks from established researchers, discussing new advancements in this area of research, characterized by the use of different theoretical approaches and tools. These talks will showcase the power of the physics-informed learning paradigm for applications from robotics to system identification to power grids. The keynote talks are complemented by two shorter talks that offer application-oriented perspectives. The workshop is aimed at a broad audience, including both junior and senior researchers, as well as practitioners.