Fire-induced damage in concrete structures is governed by strongly coupled thermo-mechanical processes that evolve across multiple time and length scales. Phase-field fracture modeling has emerged as a powerful framework for capturing complex crack patterns, spalling, and material degradation without the need for explicit crack tracking ad-hoc criteria. However, its application to fire-structure interaction poses significant modeling and computational challenges, particularly if coupled with large-scale fire simulations. This contribution presents a fully coupled fire-structure interaction methodology that integrates advanced phase-field fracture models with computational fluid dynamics. The thermal and fluid processes of the fire are simulated using the open-source Fire Dynamics Simulator (FDS), while the structural response is modeled using the phase-field method in FEniCS. In the current stage, the investigations are restricted to unreinforced concrete. Multiple phase-field approaches are employed and compared to study fire-induced damage mechanisms, including the phase-field cohesive zone model, allowing for systematic investigations of temperature-dependent material properties, crack driving forces, and different thermo-mechanical boundary conditions. A novel bi-directional coupling method is presented between fire dynamics and phase-field damage evolution, realized via the open-source coupling library preCICE[1]. Temperature fields from the fire simulation are employed as input thermal boundary conditions for the phase-field model, while the evolving damage field feeds back into the fire solver to modify the computational fluid domain. Damage-induced cracks are interpreted as evolving surface spalling and an eventual breakthrough, enabling the smoke leakage and radiative heat transfer through the concrete wall and thereby influencing the fire spread. This approach is demonstrated on a representative benchmark of fire-exposed concrete, highlighting the capabilities of the phase-field fracture models for fire-driven structural failure and discussing future challenges related to model selection, numerical efficiency, and multi-physics coupling robustness. References: [1] G. Chourdakis, K. Davis, B. Rodenberg, M. Schulte, F. Simonis, B. Uekermann et al., preCICE v2: A sustainable and user-friendly coupling library, Open Research Europe, vol.2, no.51, 2022.