The KU Leuven AI Safety Lab is inviting applications for five Visiting Fellowships 2027 for researchers interested in the large-scale risks of advanced artificial intelligence and the governance and control of AI systems. Selected fellows will spend three months in residence in Leuven, Belgium, working on a self-directed research project within one of the Lab’s research programmes.
The fellowship is designed for researchers from a range of disciplines, including philosophy, law, political science, economics, psychology, computer science and related fields. Fellows will join a multidisciplinary cohort and become part of the Lab’s day-to-day research environment.
KU Leuven AI Safety Lab Visiting Fellowship 2027 – At a Glance
- Host: KU Leuven AI Safety Lab
- Location: Leuven, Belgium
- Duration: Three months
- First cohort: March–May 2027
- Fellowship allowance: €2,000 per month
- Number of fellowships: Five
- Application deadline: 31 October 2026
- Format: Fully in residence; remote participation is not available
Research Areas
Applicants must indicate one of the Lab’s research programmes when applying.
Theories of Victory and Viable Endgames focuses on positive outcomes from advanced AI, including what a desirable future involving advanced AI could look like and the conditions required to achieve it.
Democratic Control over AI Companies examines the concentration of AI-enabled power within frontier AI companies and the legal, political and institutional mechanisms that could allow democratic societies to exercise meaningful control.
Applicants may also propose research under Other. Strong projects outside the two named programmes may be considered where there is a clear and convincing contribution to AI safety.
What the Fellowship Provides
Selected fellows will have a desk at the Lab’s Leuven offices and access to KU Leuven libraries and facilities during their three-month residence.
The fellowship provides a €2,000 monthly allowance, intended to cover accommodation and other living costs. Fellows receive formal status as visiting scholars at KU Leuven rather than employment contracts. They remain responsible for their own tax and social-security arrangements.
Fellows will receive supervision and feedback from the Lab’s core team and can present work in progress at internal seminars. They will also have opportunities to present their research at the Lab’s annual conference and summer school.
Who Can Apply?
The fellowship is open to:
- PhD candidates
- Postdoctoral researchers
- Faculty members
- Researchers based outside academia
Applicants working outside academia should demonstrate that the fellowship stay will serve a genuine research purpose.
The KU Leuven AI Safety Lab particularly encourages applications from women, minorities and other underrepresented groups.
Applicants should have a research project that can realistically be completed during the three-month fellowship and should demonstrate a strong connection between their proposed work and AI safety.
Fellowship Expectations
Fellows are expected to participate actively in the Lab’s research environment and regularly work from the Leuven office. The fellowship is therefore not suitable for researchers seeking a remote arrangement.
By the end of the fellowship, fellows are expected to complete a research paper, deposit it as a preprint on arXiv, PhilPapers or an equivalent repository, and submit it to a target publication or academic venue identified in their research plan.
Fellows must also present their work at the Lab and actively disseminate their findings after the fellowship. They are expected to contribute to the Lab’s activities, including its annual conference and/or summer school.
Co-authorship is not compulsory, although the Lab encourages collaborative research where appropriate.
Application Requirements
Applicants must submit the following materials:
Curriculum Vitae: A maximum of two pages, including a publication list.
Cover Letter: A maximum of one page explaining the applicant’s motivation, fit with the Lab and what they hope to gain from the fellowship.
Research Plan: A maximum of 1,500 words. The research plan should explain:
- The research question and why it matters for AI safety
- Proposed methodology
- Theory of change and anticipated academic, policy or field-building impact
- Intended journals, conferences and workshops for submission
- Why the project is feasible within three months
- The research programme being applied to
How Applicants Are Selected
Applications will be assessed according to the quality and originality of the proposed research, its fit with one of the Lab’s research programmes, the plausibility of its theory of change, feasibility within three months and the contribution the applicant could make to the Lab’s research community.
The call was published on 21 September 2026, with the application deadline on 31 October 2026. Shortlisted applicants are expected to be interviewed in November 2026, with successful applicants contacted by the end of November. The first cohort will be in residence from March to May 2027.
Why This Fellowship Is Valuable
The KU Leuven AI Safety Lab Visiting Fellowship offers researchers an opportunity to develop an independent AI safety project while working within a multidisciplinary academic research environment. The combination of a dedicated research period, supervision, access to KU Leuven facilities, research-community engagement and a €2,000 monthly allowance can be particularly valuable for researchers seeking to develop publishable work in an emerging and increasingly important field.
The fellowship is especially suited to applicants with a clearly defined research question, a credible methodology and a realistic plan for producing a completed paper within three months.
How to Strengthen an Application
Applicants should ensure that their research proposal is tightly scoped enough to be completed during the fellowship. The theory of change should be particularly clear, showing how the proposed research could influence academic understanding, policy or AI safety field-building.
Applicants should also demonstrate a strong fit with their selected research programme and explain why the KU Leuven AI Safety Lab is the appropriate environment for the project. A concise methodology, realistic publication targets and a clear explanation of the project’s AI safety relevance can further strengthen the application.
Application Deadline
The deadline to apply for the KU Leuven AI Safety Lab Visiting Fellowships 2027 is 31 October 2026.
Interested researchers should prepare the required documents and submit their application through the official KU Leuven AI Safety Lab application process before the deadline.
