UCL joins forces with Google DeepMind to democratise access to AI education

A new AI curriculum, designed to equip the next generation of researchers with a deep understanding of generative AI and large language models (LLMs), has been developed by UCL and Google DeepMind.

AI Research Foundations, which launches today (21 October 2025), has been designed to help address regional (in-country) barriers to accessing education and training in advanced AI research techniques. 

While available globally, the free courses have been localised for educators and learners in Africa, where access to advanced AI education has been limited.

The curriculum aims to equip learners with AI research skills that could one day help address major local and global research challenges – such as building climate resilience or tackling local diseases. Created for undergraduates to early career researchers across mathematics, computer science, physics and engineering, the courses combine responsible innovation activities, coding exercises, and an opportunity to fine-tune a language model from the ground up. 

Professor John Mitchell, Co-Director of UCL’s Centre for Engineering Education said: “This project reflects UCL’s commitment to widening access to world-leading research and innovation in AI. By working with Google DeepMind and African academic collaborators, we have co-created a curriculum that not only teaches the foundations of machine learning but also ensures that learners can apply these skills to drive positive social impact in their own communities.” 

The curriculum was conceived by a Google DeepMind team in response to a clear need for open and accessible AI research education that reflects the latest thinking in the field. Their ongoing collaboration with the Deep Learning Indaba – a movement to strengthen machine learning research in Africa – underscored the specific needs of learners and educators across the continent.

Ulrich Paquet, Research Scientist at Google DeepMind, Director of AIMS South Africa and Co-founder of Deep Learning Indaba, said: “AI is an exciting, scientifically rich discipline, holding enormous potential to help communities tackle the research challenges that matter most to them. However, without broadening participation in advanced AI education, we risk missing out on pioneering research and groundbreaking innovation. 

“We built AI Research Foundations in consultation with lecturers, students and grassroots AI communities. With practical exercises and a focus on responsible AI techniques, this curriculum gives budding AI researchers the tools to push the field forward.”

AI in Africa and commitment to free online education

Throughout the project African academic voices have consistently informed the design process, through collaboration with the African Institute for Mathematical Sciences (AIMS) and co-design workshops with educators from across the continent.

Tejumade Afonja, PhD Researcher at CISPA, Director of TRI-AI and Co-founder of AI Saturdays Lagos said: "For years, educators in Africa have relied on free and open courses to teach machine learning and AI, because creating high-quality content from scratch takes resources many communities like theirs don’t have. 

"That’s why efforts like this matter. They make it possible for us to tailor world-class learning materials to local contexts, helping more people understand the fundamentals of cutting-edge AI models, so they can meaningfully contribute to AI. What’s more, they help communities to confidently use and adapt these technologies to solve local problems and create innovations with global impact."

A dedicated team of UCL learning designers and specialists in online pedagogy have contributed, led by Matthew Seren Smith (UCL Centre for Engineering Education) and Professor Eileen Kennedy (UCL Institute of Education), alongside a group of academic developers led by Professor Pontus Stenetorp (UCL Centre for Artificial Intelligence), Dr Sebastian Schuster (UCL Honorary Research Fellow & University of Vienna) and Dr Stephen Hughes (UCL Science and Technology Studies).

The curriculum blends Google DeepMind’s leading AI research expertise with UCL’s academic excellence in engineering education and online pedagogy, AI and machine learning, and was co-developed by a dedicated team of UCL, academics, learning designers and education specialists.   

Matthew Seren Smith, Lead Learning Designer at UCL, added: “From the very beginning, our work with Google DeepMind has been driven by a shared commitment to equitable online education. By bringing together leaders from academia and industry, we’ve developed a unique programme that offers students cutting-edge technical learning in Responsible AI, while exemplifying the academic rigor and educational standards for which UCL is renowned.” 

Find out more