The project aims to enhance collaboration between air traffic controllers and AI systems in air traffic management. Its goal is to enable AI systems in human-AI teams to anticipate when and what kind of assistance is needed by controllers and respond accordingly. DIALOG is developing an AI-based digital assistant, the teamwork assistant , which uses speech recognition and understands pilot-controller exchanges to infer ATCOs' intent and goals. The assistant also employs machine learning methods to assess controllers’ workload and attention in real-time, based on voice, physiology, and behavioural data. The actions of the assistant are determined by the controllers’ current state, attention allocation, traffic situation, and overall context. The project is developing prototypes and carrying out validation exercises in close collaboration with end-users to assess the performance and usability of the proposed solutions.