The application deadline is 28 September 2025.
The fellowship is a 3 month-long research program for participants from diverse backgrounds from around the world to pursue AI Safety research from a cooperative AI perspective. The fellowship will run from January to April 2026 and kicks off with a week-long retreat and may culminate with a conference.
Participants will receive mentorship from top researchers in the field of cooperative AI, including from organisations such as Google Deepmind, the University of Oxford, and the Collective Intelligence Project (among others). Alongside this, participants will be provisioned resources for building knowledge in AI Safety, and comprehensive financial support covering their living and travel expenses while in Cape Town.
There will be opportunities for collaboration with other cohort members, a week-long retreat, workshops and access to a co-working space.
Alumni from previous iterations of similar fellowships have gone on to work in academia, join leading AI Safety labs, work for the UK AISI, or start independent research projects.
Location: In-person in Cape Town, South Africa.
Application Deadline: 28 September 2025.
Start Date: 10 January 2026.
Duration: 3 months, ending 13 April 2026.
Stipend: $1170 (R21000)/month for living expenses. Note that these are generous given the comparatively low cost of living in South Africa.
Travel Support: We cover flights to and from Cape Town.
Visas: We aren't able to cover visas, however visitors visas are easy to acquire for many countries and last up to 90 days, with relatively simple processes for extension. We can provide support in handling your visitor's visa extension process.
Compute Budget: We offer compute support through our partner organizations.
Eligibility: We welcome participants from all kinds of backgrounds, but a fundamental understanding of machine learning is required. Additional expertise in MARL, game theory, complex systems, mathematics, economics, and international governance is appreciated. We are committed to the growth of a diverse and inclusive research community and welcome applicants from underrepresented backgrounds.
Participants: There will be 6-12 fellows, with at least 30% from South African institutions. The remaining 50-70% of the cohort will be drawn from a diverse pool of international applicants.
We have identified the following tentative tracks for the fellowship.
Understand and control the collective dynamics arising from the interactions within multi-agent networks.
Develop frameworks for multi-agent systems which serve to preserve human agency, combining technical solutions with robust governance frameworks.
Research AI tools for collective decision making allowing for societally beneficial democratic engagement.
Understand and control the collective dynamics arising from the interactions within multi-agent networks.
Understand and control the collective dynamics arising from the interactions within multi-agent networks.
Develop frameworks for multi-agent systems which serve to preserve human agency, combining technical solutions with robust governance frameworks.
Research AI tools for collective decision making allowing for societally beneficial democratic engagement.
There will be a three-phase application process. The first phase will take approximately 45-60 minutes to complete. We encourage you to submit your application even if it feels unpolished—we value authenticity and substance over perfect presentation.
We will evaluate applications based on research potential, technical capability, career alignment, motivation, and personal fit. Applications are reviewed on a rolling basis with decisions made by October 6, 2025. Early submission is encouraged.
Selected candidates will be asked to complete a compensated research task ($25/hour).
Final candidates will participate in a 45-60 minute interview with program staff. There may be a second round of interview if needed.
Successful applicants will receive offers by October 29, 2025 and be matched with mentors based on research interests and compatibility. If you receive an offer from one mentor, you will work with that mentor if you choose to accept the fellowship. If you receive multiple offers from different mentors, you will have the opportunity to choose which mentor you prefer to work with. Selected applicants will have initial meetings with their matched mentors before making final decisions about the fellowship.
A description of the fellowship structure can be found here. For any further queries please contact info@cirf.com/co.za
We value inclusion and encourage applicants from diverse backgrounds. Please contact us if you require special accommodation in order to apply.
You don't need to have a specific project in mind when applying for the fellowship. Throughout the interview process, we learn more about each fellow's interests and help them find a suitable mentor and project.
Our key criterion is the potential to conduct impactful cooperative AI research. As evidence of this, we evaluate applications in terms of personal fit, technical skills and career goals. Further instructions can be found on the application form.
Each fellow will be matched with an expert mentor, who will provide feedback on their project. In addition, fellows will be guided by a research manager who will facilitate their research process and connect them to relevant resources during weekly meetings.
As AI systems become more powerful and autonomous, it is crucial to ensure they are used for good and benefit all of humanity. Cooperation is key to achieving beneficial outcomes in complex, multi-agent environments.
Designing cooperative AI systems presents unique challenges, including aligning values across different agents, managing conflicts of interest, and ensuring fairness and transparency in decision-making processes.
Through this fellowship, we aim to contribute to solving these challenges and building a future where AI systems work together cooperatively for the benefit of humanity.
South Africa has a growing AI ecosystem with diverse talent and unique challenges and opportunities for AI development. The country offers a rich environment for exploring cooperative AI research.
Through this fellowship, we aim to leverage this potential and contribute to building a strong AI research community in South Africa that can address both local and global challenges in cooperative AI.
The fellowship is a collaboration with the Cooperative AI Foundation, PIBBSS, The AI Initiative at UCT and AI Safety South Africa. We build on PIBBSS fellowship methodology with CAIF research oversight. This initiative serves as both a talent pipeline and research direction-setting mechanism for the University of Cape Town's emerging AI Safety Hub.
A new charitable entity, backed by a $15 million philanthropic commitment from Macroscopic Ventures. CAIF's mission is to support research that will improve the cooperative intelligence of advanced AI for the benefit of all.
Learn more hereFacilitates knowledge transfer from fields studying intelligence in natural systems toward building human-aligned AI.
Learn more hereProvides a rich ecosystem that enables researchers and students to do excellent AI research which will create tools and opportunities that facilitate the transition to a just society.
Learn more hereA capacity building organisation focused on developing skills and community in South Africa through events, courses, and co-working at AI Safety Cape Town.
Learn more here
The fellowship is a collaboration with the Cooperative AI Foundation, PIBBSS, The AI Initiative at UCT and AI Safety South Africa. We build on PIBBSS fellowship methodology with CAIF research oversight. This initiative serves as both a talent pipeline and research direction-setting mechanism for the University of Cape Town's emerging AI Safety Hub.
A new charitable entity, backed by a $15 million philanthropic commitment from Macroscopic Ventures. CAIF's mission is to support research that will improve the cooperative intelligence of advanced AI for the benefit of all.
Learn more hereFacilitates knowledge transfer from fields studying intelligence in natural systems toward building human-aligned AI.
Learn more hereProvides a rich ecosystem that enables researchers and students to do excellent AI research which will create tools and opportunities that facilitate the transition to a just society.
Learn more hereA capacity building organisation focused on developing skills and community in South Africa through events, courses, and co-working at AI Safety Cape Town.
Learn more here
If you have further questions, please reach out to us at admin@aisafetysa.com.
