The interface between gravity and quantum theory has become a lively field of research, substantially driven by the developments of quantum optics technologies. One of the most ambitious experimental goals is the preparation of a massive source in a quantum superposition of spatial locations, enabling tests of gravity in genuinely quantum regimes. Such experiments raise crucial theoretical questions: what insight could they provide on the (quantum) nature of gravity, and which tools and techniques are best suited to interpret them? Information-theoretic methods have proved very useful to analyse these scenarios, with entanglement and the LOCC theorem playing a central role. While these methods are naturally suited to the experimental platforms under consideration, their application to gravity is complicated by the gauge-theoretic nature of the field, which challenges standard notions of subsystems, locality, and information. 

In this talk, I will argue that concepts and tools of quantum information should be generalised to accommodate gauge theories and gravity. After reviewing the role played by information-theoretic arguments in current proposals, I will discuss how some of these notions can be reformulated in a gauge-invariant framework. This perspective provides a route toward extending the methodology of quantum information to gravitational scenarios.