Resumen
AI is getting more and more embedded into our daily life, ranging from AI assistants to recommendation, decision-support,
smart and autonomous systems. It is therefore crucial for AI-based systems to align with us humans and our values and
preferences, so we can make the most of AI minimising any risks associated to its use. For this, AI needs to be humancentred,
that is, it needs to: 1) behave ethically and respect our privacy; 2) reason about, learn and adapt to our preferences
and values; and 3) explain what it does and why, in a way that humans can understand and act upon regardless of their AI
literacy. HAIFEL will pave the way towards AI that is more human-centred through novel AI models for privacy-respecting
and ethical AI that consider human values and user preferences for decision-making, and that are able to explain how they
do it.
