Iliad Fellowships Fall 2026 (Fully Funded AI Alignment Research Fellowship)

Applications are now open for the Iliad Fellowships Fall 2026, a fully funded three-month research fellowship designed for mathematicians, theoretical computer scientists, physicists, and researchers interested in advancing the field of AI alignment. Organized by Iliad in collaboration with PrincInt, the fellowship provides participants with the opportunity to conduct mentored research on foundational mathematical questions related to artificial intelligence safety while working alongside leading researchers at the London Initiative for Safe AI (LISA).

As artificial intelligence systems become increasingly capable, ensuring that these systems remain safe, reliable, and aligned with human values has become one of the most important scientific challenges of the modern era. The Iliad Fellowship was established to attract talented researchers from mathematics and related disciplines into AI alignment research by providing funding, mentorship, and an intellectually stimulating environment where fellows can explore fundamental theoretical questions.

The fellowship emphasizes rigorous mathematical thinking rather than applied software engineering. Participants will investigate topics that contribute to a deeper understanding of AI systems while developing research proposals that can serve as the foundation for long-term careers in AI safety and alignment research. Through close collaboration with mentors and fellow researchers, participants will gain valuable experience in one of the fastest-growing areas of artificial intelligence research.

About the Fellowship

The Iliad Fellowship is an intensive, full-time research residency that enables outstanding researchers to transition into or deepen their expertise in AI alignment. Rather than offering classroom-style instruction alone, the programme focuses on independent research supported by experienced mentors with expertise in mathematics and AI alignment.

During the three-month residency, fellows work on carefully selected research projects that address foundational problems in AI safety. Participants also attend seminars, engage in technical discussions, collaborate with peers, and receive continuous feedback from supervisors throughout the programme.

A major objective of the fellowship is to help participants produce a detailed research proposal capable of supporting future academic work or funding applications. By the conclusion of the programme, fellows are expected to have developed both a strong research direction and meaningful connections within the international AI alignment community.

Fellowship Benefits

Successful applicants will receive:

• Monthly travel and housing allowance of USD 6,000.
• Full-time participation in a three-month research fellowship.
• Mentorship from experienced AI alignment researchers.
• Office space at the London Initiative for Safe AI (LISA).
• Opportunities to attend technical seminars and workshops.
• Collaboration with an international community of researchers.
• Guidance in developing competitive research proposals.
• Exposure to future research funding opportunities.
• Networking with experts working in AI safety and alignment.
• Support for pursuing long-term careers in AI alignment research.

Research Areas

Participants may work on research topics including:

• Mathematical foundations of AI alignment.
• Solomonoff induction.
• Simplicity and degeneracy.
• Universal prediction.
• Causality.
• Computational mechanics.
• Algorithmic statistics.
• Learning theory.
• Training dynamics.
• Sparse coding.
• Lossy compression.
• Abstraction and coarse-graining.
• AI interpretability.
• Reward learning.
• Value learning.
• Debate and scalable oversight.
• Other mathematical problems relevant to AI alignment.

Applicants are also encouraged to propose original research ideas aligned with the programme's objectives.

Eligibility Criteria

Applicants should satisfy the following requirements:

• Be a PhD student, postdoctoral researcher, experienced academic, or independent researcher.
• Possess a strong background in mathematics, theoretical physics, theoretical computer science, or a closely related discipline.
• Demonstrate excellent analytical and mathematical reasoning skills.
• Have an interest in AI alignment and AI safety research.
• Be available to participate full-time throughout the fellowship period.
• Be willing to conduct research either in London or under approved exceptional arrangements where applicable.

The fellowship welcomes applicants from all countries. Formal academic qualifications are preferred, but equivalent research experience may also be considered.

Required Application Materials

Applicants are generally expected to prepare:

• Completed online application form.
• Curriculum Vitae (CV).
• Academic and research background information.
• Description of research interests.
• Information demonstrating mathematical or theoretical expertise.
• Any additional supporting materials requested during the application process.

Applicants should present evidence of their research ability and explain why they are interested in contributing to AI alignment research.

Selection Criteria

Applications will be assessed based on:

• Mathematical ability.
• Research experience.
• Academic excellence.
• Analytical thinking.
• Potential to contribute to AI alignment research.
• Fit with available mentors and research projects.
• Motivation for participating in the fellowship.

The programme is highly competitive, and applicants are selected based on their demonstrated research potential and ability to contribute to foundational AI safety research.

How to Apply

To apply for the Iliad Fellowships Fall 2026:

1. Visit the official Iliad Fellowship application portal.
2. Review all eligibility requirements carefully.
3. Complete the online application form.
4. Upload all requested supporting documents.
5. Submit your application before the application deadline.
6. Await communication regarding interviews and selection outcomes.

Applicants are encouraged to apply as early as possible to maximize their chances of selection.