Artificial Intelligence is rapidly transforming industries, economies, and societies around the world. As AI systems become increasingly capable, ensuring they remain safe, reliable, and aligned with human values has become one of the most important scientific challenges of our time. To address these pressing concerns, AIXI Labs is inviting applications for its Research Fellowship, a prestigious opportunity designed for talented researchers interested in advancing the field of AI safety through rigorous theoretical and empirical research.

The AIXI Labs Research Fellowship offers participants the opportunity to work alongside some of the world’s leading AI safety researchers while receiving a monthly stipend ranging from £2,000 to £3,500, a dedicated machine learning compute budget, close mentorship, and the flexibility to work fully remotely from anywhere in the world or from the organization’s headquarters in London, United Kingdom.

Applications are accepted on a rolling basis, making this an excellent opportunity for aspiring AI researchers to join an internationally recognized research community focused on reducing existential risks from advanced artificial intelligence.

About AIXI Labs

AIXI Labs is a London-based nonprofit research organization dedicated to advancing AI safety through the lens of Algorithmic Information Theory (AIT).

Unlike many organizations that either focus exclusively on abstract mathematical models or highly practical machine learning systems that are difficult to analyze theoretically, AIXI Labs occupies a unique position by combining both rigorous theory and practical applications.

The organization’s research focuses on developing methods that are:

General enough to describe highly capable intelligent agents
Mathematically rigorous enough to produce provable safety guarantees
Applicable to modern Large Language Models (LLMs) and AI agents

The research agenda is built around AIXI, a well-known theoretical model of unbounded artificial superintelligence developed to study idealized intelligent agents. Researchers at AIXI Labs aim to translate theoretical breakthroughs into practical safety techniques applicable to modern AI systems.

About the AIXI Labs Research Fellowship

The AIXI Labs Research Fellowship is a quarterly rolling fellowship program created for researchers who want to make meaningful contributions to AI safety research.

Participants spend approximately one academic quarter working on a focused research project under the guidance of experienced mentors while collaborating with a growing international AI safety community.

The fellowship is designed to support independent researchers who wish to explore foundational questions while maintaining enough flexibility to pursue topics aligned with their own expertise and interests.

Fellowship Highlights

Selected fellows will receive:

Monthly stipend between £2,000 and £3,500
Dedicated machine learning compute budget
Close mentorship from leading AI researchers
Flexible remote work from anywhere in the world
Optional opportunity to work from London headquarters
Opportunity to publish research papers and technical articles
Access to one of the world’s emerging AI safety research communities
Potential renewal for exceptional fellows

Fellowship Duration

The fellowship runs on a quarterly basis, with each fellowship lasting approximately one academic quarter.

Applications are reviewed continuously throughout the year under a rolling admissions process.

Exceptional participants may be invited to extend or renew their fellowship.

Fellowship Location

Applicants may choose either of the following work arrangements:

Fully Remote (Worldwide)
London, United Kingdom

Researchers can participate from any country while collaborating virtually with mentors and fellow researchers.

Financial Support

Successful fellows receive substantial financial support throughout the program.

This includes:

Monthly stipend between £2,000 and £3,500
Dedicated compute budget for AI and machine learning experiments

The compute budget enables fellows to perform computationally intensive machine learning research without bearing personal infrastructure costs.

Research Mentorship

One of the strongest aspects of the fellowship is direct mentorship from AIXI Labs’ core research leadership.

Fellows work closely with:

Cole Wyeth — Founder & Executive Director
Marcus Hutter — Research Director
Aram Ebtekar — Founding Research Scientist

This close supervision provides participants with valuable guidance while allowing them to maintain ownership over their research projects.

Research Focus Areas

The fellowship supports projects across a broad range of AI safety topics grounded in algorithmic information theory.

Potential research directions include:

1. AI Safety Theory

Developing mathematical foundations for understanding safe intelligent agents.

2. Algorithmic Information Theory

Applying information-theoretic approaches to model intelligence and alignment.

3. Reinforcement Learning

Exploring safer reinforcement learning systems and decision-making agents.

4. Large Language Models

Studying the safety, behavior, and alignment of LLM-based agents.

5. AI Alignment

Investigating methods for aligning advanced AI systems with human goals.

6. Empirical AI Safety

Testing theoretical ideas using practical machine learning experiments.

Responsibilities of Fellows

Participants are expected to pursue a self-directed research agenda aligned with AIXI Labs’ mission.

Primary responsibilities include:

Conduct original AI safety research
Design and execute independent research projects
Align research with algorithmic-information-theoretic approaches
Perform empirical experiments where applicable
Produce high-quality written research outputs
Publish papers, technical reports, or blog posts
Communicate findings to the AI safety and machine learning communities

Researchers are encouraged to tailor their projects according to their own strengths and academic interests.

Minimum Eligibility Requirements

Applicants should demonstrate strong academic preparation in quantitative disciplines.

Minimum qualifications include:

Research background equivalent to a current PhD student
Alternatively, applicants with master’s-level mathematical expertise combined with substantial industry or applied research experience may also qualify for primarily empirical projects
Strong knowledge of:
- Probability
- Statistics
- Information Theory
- Reinforcement Learning
Excellent written and verbal communication skills
Strong interest in reducing existential risks associated with advanced AI
Enthusiasm for AI safety research

Importantly, prior experience with AIXI or Algorithmic Information Theory is not required.

Preferred Qualifications

While not mandatory, competitive applicants may possess one or more of the following:

Publications at leading AI conferences such as:
- NeurIPS
- ICML
- ICLR
- ALT
- COLT
Background in:
- Algorithmic Information Theory
- Computability Theory
- Learning Theory
- Theoretical Computer Science
Experience:
- Fine-tuning Large Language Models
- Evaluating LLMs
- Red-teaming AI systems
- Reinforcement Learning agents
Previous involvement with:
- UAI community
- AIXI community
- Related reading groups

AIXI Labs emphasizes the quality and impact of an applicant’s strongest research contributions over the sheer number of publications.

Click HERE to read more.

Why Apply for the AIXI Labs Research Fellowship?

This fellowship offers an exceptional opportunity for researchers seeking to work on one of the most important scientific challenges of the coming decades.

Benefits include:

Competitive monthly financial support
Fully remote participation
Flexible project selection
Direct mentorship from internationally respected AI researchers
Dedicated compute resources
Opportunity to contribute to frontier AI safety research
Collaboration with an international research community
Potential fellowship renewal

Rolling Applications

Unlike programs with fixed annual deadlines, the AIXI Labs Research Fellowship reviews applications throughout the year.

This rolling admissions model allows qualified researchers to apply whenever they are ready, increasing accessibility for applicants worldwide.

Previous applicants remain eligible to apply for future fellowship rounds.

Deadline: There is no fixed application deadline. Applications are reviewed on a rolling basis throughout the year.

Who Should Apply?

The fellowship is particularly suited for:

PhD students
Early-career AI researchers
Machine learning researchers
Reinforcement learning researchers
Computer scientists
Mathematicians
Statisticians
AI alignment researchers
Theoretical computer scientists
Researchers interested in AI existential risk

Candidates with strong mathematical backgrounds and a genuine passion for advancing AI safety will benefit most from this opportunity.

APPLY HERE

Conclusion

The AIXI Labs Research Fellowship represents an outstanding opportunity for researchers who want to contribute to one of the world’s most significant scientific and technological challenges: ensuring the safe development of advanced artificial intelligence.

By combining generous financial support, flexible remote participation, dedicated mentorship, computational resources, and access to a vibrant AI safety research community, the fellowship enables participants to pursue cutting-edge research with real-world impact. Whether your interests lie in theoretical AI, reinforcement learning, algorithmic information theory, or large language model safety, this fellowship provides an ideal environment to deepen your expertise while contributing to the future of responsible AI.

Click HERE for more fellowships.

AIXI Labs Research Fellowship 2026: Fully Remote AI Safety Fellowship with Monthly Stipend of £2,000–£3,500 (Worldwide Applications Open) ·£3,500 (Worldwide Applications Open)