In the realm of artificial intelligence, tackling the intricate challenge of user mistrust is paramount. This scepticism is particularly prominent in the domain of sequential decision-making problems. These AI applications involve intricate sequences of actions, where transparency and tailored solutions are often lacking, leading to user misinterpretations, hesitations, and an overall lack of acceptance. In this context, the EU-funded PEER project aims to reshape the landscape of human-AI collaboration. PEER’s pioneering approach entails placing users at the heart of the AI process, fostering bidirectional communication and mutual learning. Bridging social sciences and AI, PEER crafts new engagement methods, devises AI planning techniques for dynamic scenarios, introduces an AI acceptance index, and tests these concepts.