The AIXI Labs Research Fellowship 2027 is a paid research opportunity for researchers interested in AI safety, algorithmic information theory, reinforcement learning, and intelligent agents. Hosted by AIXI Labs, a London-based non-profit research organisation, the fellowship allows selected researchers to work on a focused, self-directed project with close mentorship from the lab’s research team.
The fellowship is offered on a quarterly basis with rolling applications, meaning there is no single annual deadline. Fellows can participate remotely from anywhere in the world or work from the AIXI Labs office in London. Successful researchers receive a monthly stipend of £2,000 to £3,500, depending on experience, along with a compute budget for machine-learning research.
AIXI Labs focuses on AI safety research through the ideas of algorithmic information theory, universal induction, and the AIXI model of artificial intelligence. The fellowship is intended for researchers who want to investigate foundational questions about intelligence and contribute to research related to the safety of advanced AI systems.
The programme generally lasts for about one quarter, or roughly three months. Fellows work on a defined research project and receive guidance from members of the AIXI Labs team. Exceptional fellows may be considered for renewal, although applicants should plan around a single fellowship term.
| Fellowship Details | Information |
|---|---|
| Fellowship Name | AIXI Labs Research Fellowship 2027 |
| Host Organisation | AIXI Labs |
| Location | London, UK or Remote |
| Duration | Approximately 3 months |
| Monthly Stipend | £2,000–£3,500 |
| Research Areas | AI Safety, Algorithmic Information Theory, Reinforcement Learning |
| Eligible Nationalities | Worldwide |
| Application Deadline | Rolling, reviewed quarterly |
| Participation | Remote or London-based |
| Funding | Paid fellowship + research compute budget |
Selected fellows receive a monthly stipend ranging from £2,000 to £3,500, with the amount depending on the applicant’s experience and background. In addition, AIXI Labs provides a compute budget for machine-learning research, which can support experiments involving modern AI systems.
The fellowship also provides close research mentorship and an opportunity to develop work that can contribute to the wider AI safety and machine-learning research communities. Fellows may produce research papers, technical work, blog posts, or other research outputs during their fellowship.
Applicants should note that the fellowship is not presented as a university scholarship. It is a paid research fellowship, and relocation or visa sponsorship is not advertised as part of the standard funding package.
The AIXI Labs fellowship is particularly focused on research at the intersection of theoretical foundations of intelligence and practical AI safety. Potential areas include:
Algorithmic Information Theory
Universal Induction
AI Safety
Reinforcement Learning
Intelligent Agent Systems
Large Language Model Agents
Machine Learning
Learning Theory
Theoretical Computer Science
Agent Foundations
The programme can accommodate both theoretical and empirical research projects. Applicants should develop a project that fits AIXI Labs’ research agenda and can realistically make meaningful progress within a fellowship term.
The fellowship is open to researchers worldwide, and remote participation is available. AIXI Labs generally looks for candidates with research experience comparable to that of a current PhD student.
For theoretical projects, applicants are expected to have substantial research experience at approximately PhD level. For empirical projects, the fellowship can also consider applicants with Master’s-level mathematics and significant industry or applied research experience.
Applicants should have a strong foundation in subjects such as:
Probability
Statistics
Information theory
Reinforcement learning
Machine learning
Mathematics
Theoretical computer science
Strong communication skills and genuine interest in AI safety are also important. Publications at venues such as NeurIPS, ICML, ICLR, ALT, or COLT can strengthen an application, while experience with algorithmic information theory, LLM agents, reinforcement-learning systems, or the UAI/AIXI research community is considered advantageous rather than universally mandatory.
Applicants should first study the research agenda of AIXI Labs and identify a research question that aligns with its work. Because the fellowship is self-directed, the proposed project is an important part of the application.
Candidates should prepare a strong research proposal explaining the problem they want to investigate, their proposed approach, and why the project fits the AIXI Labs research agenda. A current CV highlighting research experience, technical skills, publications, projects, and relevant industry work should also be prepared.
Applicants should demonstrate their mathematical and technical background through research papers, preprints, theses, technical writing, code, or other evidence of research ability. They should also explain their motivation for working on AI safety and indicate their availability for the fellowship period.
Applications are submitted through the AIXI Labs application process and are reviewed on a rolling basis.
The AIXI Labs Research Fellowship 2027 does not have a fixed annual deadline. Applications are accepted on a rolling basis and reviewed quarterly. Applicants can therefore submit their applications when their research proposal and supporting materials are ready.
The programme operates in quarterly research periods, so applicants should indicate when they would be available to participate. Since the fellowship involves close mentorship within a relatively small research team, applicants should also be realistic about the scope of their proposed project.
The AIXI Labs Research Fellowship provides an opportunity for experienced researchers to spend a focused period working on foundational AI safety questions while receiving financial support and research mentorship.
The combination of a £2,000–£3,500 monthly stipend, research compute budget, international eligibility, remote participation, and access to researchers working on algorithmic information theory makes the programme relevant to researchers with strong mathematical, theoretical, or machine-learning backgrounds.
For researchers outside the UK, remote participation also provides an alternative to relocating to London. Applicants interested in an in-person experience can instead explore participation from the AIXI Labs London office.
The AIXI Labs Research Fellowship 2027 is a paid research opportunity for researchers working on AI safety and related technical areas. Fellows can spend approximately three months pursuing a focused research project while receiving a monthly stipend of up to £3,500 and access to computing resources.
The fellowship is particularly suited to researchers with strong foundations in mathematics, information theory, reinforcement learning, machine learning, or theoretical computer science. Since applications are reviewed on a rolling basis, candidates should focus on preparing a clear and technically credible research proposal before applying. Apply Now