Part-time, fully remote. 13 weeks. September through December 2026. The BASE Fellowship exists specifically to put Black researchers, practitioners, and students into the rooms where AI safety decisions get made, and applications for the Fall 2026 cohort close June 16, 2026. Less than four weeks away.
SEE ALSO: Rowland Fellowship at Harvard 2027: Funded Lab, Five Years, Open Now
About BASE
Black in AI Safety and Ethics (BASE) is a global community built around one explicit goal: increasing Black representation and leadership in AI safety, governance, and security. The organisation connects, trains, and supports Black researchers and professionals to shape how AI systems are developed and governed. The concern driving it is concrete: as AI becomes more consequential, the Black diaspora risks being written out of the decisions that affect it most.
BASE runs the fellowship as its primary programme for developing the next generation of researchers in this space.
What the BASE Fellowship involves
The Fall 2026 cohort runs 13 weeks, from September to December 2026. The programme is part-time and fully remote, designed to fit around existing jobs, studies, or other commitments.
Fellows are placed in one of two track types, technical or non-technical, with specific mentor-led research projects within each track. The structure is mentorship-first: Fellows work directly with expert mentors on real research, not synthetic exercises.
What the programme provides:
- Structured expert mentorship throughout the 13 weeks
- Foundational training in AI safety, security, or governance depending on your track
- Hands-on research experience on active projects
- Access to a global network of Black professionals and researchers in AI
Stipend or compensation: not disclosed in the programme listing - confirm if the fellowship is paid before applying, especially if you are weighing it against other commitments.
Eligibility
BASE is clear about who the fellowship is for: Black researchers, practitioners, and students. The diaspora identification question in the application is voluntary self-identification, but the organisation's mission and selection criteria are explicitly built around increasing Black representation.
There is no stated geographic restriction. The application form accepts international city and country entries and asks for a time zone preference, which suggests BASE accepts applicants from outside the US.
No minimum degree requirement is stated. The application accepts candidates from any education level, from current students through working professionals.
Both technical and non-technical backgrounds are accepted. The selection process routes applicants to the appropriate track and coding or work test accordingly.
The selection process
BASE runs a two-phase process:
Phase 1: Application review. Applicants who pass the initial review receive either a coding test (technical track) or a work test (non-technical track). Candidates who pass the test stage then apply to specific mentor projects aligned with their chosen track.
Phase 2: Mentor review. Mentors review applicants who selected their track and passed. Phase 1. Mentors choose their mentees. Final offers are issued by the BASE team.
The process is competitive at each stage. Passing the initial review does not guarantee a mentor match.
SEE ALSO: University of Melbourne Graduate Research Scholarship 2026
Application materials
The application is submitted through the BASE Airtable form. You will need:
- Resume or CV
- LinkedIn profile (optional)
- Google Scholar profile, ORCID ID, or GitHub/personal website (all optional)
- Your current city, country, and time zone
- Written responses on: your motivation for the track you selected, your professional background and expertise, your research interests and goals within BASE, how your experience prepares you for your chosen track, and a specific example of a research project you led or contributed to significantly
You will also be asked about your weekly availability and how you plan to balance the programme with existing obligations. Answer this honestly - the programme is part-time but requires sustained commitment over 13 weeks.
Preparation time: The written responses are substantive. Budget at least a week to draft and refine them. The research contribution question in particular requires a specific example with your role and measurable impact, not a general description of your work.
Is this open to Nigerians and African diaspora applicants?
Yes - and BASE is arguably one of the few AI safety programmes where African and Black diaspora applicants are not an afterthought. The fellowship exists precisely because Black professionals are underrepresented in AI safety, governance, and security globally.
For Nigerian applicants: the deadline is June 16 at 11:59PM. Anywhere on Earth (AOE) means you have until June 16 at 11:59PM in the latest time zone on Earth, which is ahead of Nigerian time (WAT). You are safely within the window if you submit on June 16 in Nigerian local time.
SEE ALSO: Liverpool Master's Scholarships 2026: Up to £7,000 for International Students
How to apply
- Go to the official BASE application form
- Complete all required fields - basic information, education and work background, track selection, and written responses
- Upload your resume or CV
- Submit by June 16, 2026, at 11:59PM. Anywhere on Earth (AOE)
No reference letters are required at the application stage. If you pass the initial review, you will receive a coding or work test before any mentor interaction.
Common mistakes to avoid:
- Writing generic answers to the research contribution question - name your actual role, the specific project, and what changed because of your work
- Choosing a track based on what sounds impressive rather than where your actual experience sits mentors review track-matched applications and will notice a mismatch
- Underestimating the availability question 13 weeks of part-time research requires real weekly hours, not occasional check-ins
GizPulse verdict
The fellowship is not a diversity add-on to a standard programme - it is built from the ground up around the specific goal of Black representation and leadership in AI safety, governance, and security. That framing matters for how the mentorship works, who the cohort is, and what the network looks like after you finish.
The unknowns are real. Compensation is not confirmed. The named tracks are not listed publicly before you apply. The number of fellows selected per cohort is not disclosed. You are applying with less information than most programmes make available upfront.
Best suited for Black researchers, students, or practitioners who are already curious about AI safety, governance, or security and want structured research experience and expert mentorship in the field - not an introduction to AI generally.
