Building the Foundations of AI Evaluation
Artificial Intelligence is advancing faster than our ability to fully understand it.
That’s why the International Programme on AI Evaluation: Capabilities & Safety was created - to train the next generation of researchers and professionals who will help the world test, measure, and make AI accountable.
In this series of short interviews, our Academic Directors share what AI Evaluation means in practice, why it matters now, and how we’re building a global community to advance this new science.
Why AI Evaluation Matters
Before we can make AI safe, we have to understand it.
In this opening video, José Hernández-Orallo and Pablo Moreno explain why AI Evaluation has become a critical field — and why the world urgently needs experts who can assess what these systems can (and can’t) do.
What Does AI Evaluation Mean in Practice?
AI Evaluation isn’t just about benchmarks and numbers — it’s about understanding how AI behaves in the real world.
Here, José breaks down what it means to evaluate AI systems through modelling, inference, and insights from psychology and complex systems.
A New Discipline for a New Era
Our mission is to establish AI Evaluation as a rigorous academic field — one that will guide how the world understands and governs AI systems for generations to come.
“Maybe in a few years, people will say, I’m an AI evaluator — that’s what I do for a living.”
The Impact We Hope Graduates Will Have
Graduates of this programme will carry these lessons into their careers — helping shape policy, improve safety standards, and ensure that AI remains interpretable, ethical, and beneficial to society.
A Global Community with a Shared Language 🌍
Our goal goes beyond training; we’re building a global community of evaluators who share methods, insights, and a common language.
This community will stay connected long after the programme ends, shaping how AI is tested and trusted worldwide.
Who We’re Looking For 🎓
Who joins this programme?
Curious minds from every discipline — from computer science to policy — united by a drive to understand AI and make it beneficial for humanity.
Capstone Week 🇪🇸 : Where It All Comes Together
After months of online learning, participants gather in Valencia for an immersive week of collaboration, workshops, and real-world projects.
This is where ideas become experiments — and where the future of AI Evaluation begins to take shape.
Final Thoughts
This programme is more than a course: it’s the beginning of a field, a community, and a shared commitment to making AI systems transparent, reliable, and aligned with human values.