Student Research Programme Spring 2027
The MAIO Student Research Programme (SRP) is a hands-on AI research programme for Malaysian students. Selected students receive mentorship from a research advisor over 12 weeks to turn their AI research idea into a real project. The programme runs entirely online, and there is no in-person component. Proposal submissions for the Spring 2027 edition open on Apr 4, 2027.
Goal
The goal of this programme is to give Malaysian students a structured pathway into AI research, extending upon the existing MAIO Training and Selection Programme.
Timeline - Spring 2027 edition
MAIO Student Research Programme staff may reach out to you for clarification on your proposal contents as part of the proposal review procedure. Thus, you are recommended to prepare and submit your proposals early.
Eligibility
The Spring 2027 edition is open to all students who have attended the MAIO Training and Selection Programme in 2024, 2025, or 2026. Only 3–5 students will be selected for the first cohort of this programme.
Fees
Participation is free of charge to students.
Proposal selection criteria
Research proposals are evaluated directly by programme staff based on the following:
- Impact: the author should be able to demonstrate why the outcome of the research proposal is meaningful
- Execution: the research proposal is scoped to be completable and achievable within the timeline of the student research programme
- Follows the scientific method: the research proposal contains a testable hypothesis, where the hypothesis can be validated or invalidated based on further experimental data
- Grounded: the research proposal should contain a review of existing relevant work, and address why the proposed work matters in context to the existing work
- Practical or frontier-relevant: the outcome of the research proposal is either useful to a user persona, or engages the fast moving edge of the field
Research focus areas
A proposal must be rooted in the field of artificial intelligence. Any area within AI is accepted, however the programme will preferentially accept topics within the following focus areas:
- Representations and latent space: how do models encode concepts, facts, and behaviours internally, and what we can do with that? Example keywords: activation steering, interpretability, how facts get memorized
- Low-compute regime: how far can you train or run small models on limited hardware? Example keywords: small models, distillation, quantization, using a large model to guide a small one
- Low-data regime: what can you achieve with less than one hundred samples instead of a million? Example keywords: few-shot learning, data augmentation, synthetic data
- AI slop and misinformation: the internet is filling up with AI-generated content at scale with varying levels of trust and verification. How do we mitigate its impact on society? Example keywords: detectors, evaluations, deduplication
- AI in education: modern education is heavily disrupted by AI. Teachers cannot prevent students from using AI to write essays or do online assessments without supervision. Proctoring does not scale to every task, and the impact of wrongfully classifying a student submission as cheating is high. How can we tell what a student actually produced work on their own without infringing upon student privacy?
The conditional acceptance process
We believe students deserve support in shaping their idea, not just a verdict on their first draft. Thus, instead of a single accept/reject decision, we use a conditional acceptance process. This mirrors standard practice in the scientific world, where proposals go through rounds of edits before they are accepted. Programme staff will work with you directly to refine your proposal until it is ready.
Please note that proposals which have received a conditional acceptance may still not proceed into the Student Research Programme if the proposal is not sufficiently strengthened during the Proposal Refinement Window.
Intellectual property policy
All ideas and code developed under the Student Research Programme do not belong to MAIO, but to the participating students. MAIO behaves in a purely advisory capacity and does not claim ownership, copyright, or authorship over your work.
Frequently asked questions
Q: Can I submit proposals for multiple topics?
A: Yes, but we will ask you to only work on a single topic during the SRP. Please do not optimize for maximizing the number of proposals submitted, as splitting your time will dilute the quality of your research proposals.
Q: How much time should I dedicate to research if I get selected?
A: We estimate it to be 8–12h a week. Your mileage may vary based on your research topic.
Enquiries
Email us at hello <at> aiolympiad.my.