A single application can now open doors to between fifty and a hundred research institutions across Europe, under a scheme designed to keep top machine learning talent working within the continent rather than departing for labs in North America or Asia. The ELLIS PhD & Postdoc Program, run by the European Laboratory for Learning and Intelligent Systems, has grown into one of the more structurally ambitious efforts to coordinate doctoral training in artificial intelligence across national borders.
How the program is structured
The model rests on dual supervision: each student is paired with advisors in two different European countries, an arrangement intended to expose researchers to distinct institutional cultures, methodologies, and industry contacts before they settle into a single research trajectory. Candidates choose among three tracks of study, and the application process itself has been centralised, meaning prospective students no longer need to apply separately to dozens of universities and institutes. That administrative simplification matters more than it might first appear. Doctoral recruitment in AI is fiercely competitive, and researchers who might otherwise be discouraged by fragmented national systems, and inconsistent funding timelines, gain a clearer route into the field.
A defining feature is the six-month research exchange, which places students in a second lab environment partway through their doctorate. Combined with cross-sector collaboration involving companies alongside academic partners, the structure mirrors a broader shift in AI research, where the boundary between industrial labs and university departments has grown less distinct. Recent backing from G-Research, a quantitative research and technology firm, reflects that trend: private-sector investment in doctoral pipelines has become a recognisable feature of how AI talent is now cultivated in Europe.
What the network offers beyond supervision
Beyond formal academic mentorship, the program functions as a recurring calendar of events meant to build a durable research community rather than a loose affiliation of scholarships. Winter and summer schools, doctoral symposia, and alumni gatherings, hosted by ELLIS Units in cities including Amsterdam, Lisbon, Warsaw, Tübingen, and Manchester, give students recurring opportunities to present work and meet collaborators outside their home institution. One PhD student, Wafaa Mohammed at the University of Amsterdam, has described how a winter school, a Lisbon exchange, and a symposium presentation combined to shape her research trajectory rather than functioning as isolated add-ons.
The governance structure behind this network includes Fellows and Scholars, senior researchers at institutions such as Oxford, ETH Zurich, Technion, and CISPA Helmholtz Center, who hold advising privileges and can take on ELLIS PhD candidates directly. This layer of accountability, tied to named institutions rather than a loose consortium, gives the program a degree of institutional weight that has helped it expand its call for new advisors and proposals for future schools well into 2027 and 2028.
Why this matters for European AI policy
Europe has struggled for years with a familiar pattern: strong doctoral training followed by talent migration toward better-resourced labs abroad. Programs built around joint supervision, funded exchanges, and cross-institutional recognition attempt to address that directly, by making it easier for a student to build a career, and a professional network, without leaving the continent. Whether such structures meaningfully change long-term retention is a question that will take years to answer, but the design itself signals how seriously European research bodies now treat coordinated doctoral training as an instrument of competitiveness, not merely an academic convenience.