CIP 2026 Research Fellowship on AI, Public Values and Democratic Governance

The Collective Intelligence Project has opened applications for its 2026 Research Fellowship programme, giving selected researchers access to exclusive AI governance datasets and a stipend of $9,000 over six months.

The fellowship is designed for researchers interested in studying the relationship between artificial intelligence, democratic systems, public opinion, and governance frameworks. CIP said the pilot cohort will select only three fellows globally.

According to the organisation, the programme will run remotely and is non-residential, allowing researchers from different countries to participate without relocation.

Applications close on May 15, 2026.

What the CIP Research Fellowship offers

The fellowship will last for six months with an expected commitment of between five and 10 hours weekly. Fellows will also participate in bi-weekly one-hour virtual lab meetings.

Successful applicants will receive:

• A fixed stipend of $9,000 paid in three milestone-based tranches
• Access to exclusive datasets and internal research tools
• Opportunities for co-authorship on CIP Working Papers
• Public visibility through CIP communication channels
• Collaboration opportunities with other fellows
• Direct involvement in future CIP research initiatives

CIP said the programme is aimed at supporting original and publishable work focused on AI governance and public participation in AI systems.

Datasets available to selected fellows

Researchers selected for the programme will gain internal access to three major CIP projects.

The first is “Global Dialogues,” a recurring multi-country survey project tracking public attitudes toward AI in more than 70 countries. The dataset reportedly includes binary votes, demographic segmentation, pairwise comparisons, and open-ended responses collected across more than seven completed rounds.

Another dataset is “Collective Generation,” described as CIP’s deliberative framework for building AI evaluation systems through consensus-building processes.

The third project, “Weval,” is an open AI evaluation platform containing more than 100 community-contributed evaluation blueprints across over 112 AI model configurations.

According to CIP, the datasets provide opportunities for researchers to explore how different populations think about AI governance, ethics, regulation, and institutional trust.

Research themes for the fellowship

The organisation said each fellow will work from a distinct thematic perspective to reduce overlap and encourage complementary outputs.

One thematic area focuses on “Democratic and Comparative” research. This includes studying public attitudes toward AI governance across countries, examining cultural influences, and analysing how consensus or disagreement forms among different populations.

Another focus area is “Communication and Deliberation,” which centres on how people discuss, negotiate, and explain AI-related values during deliberative sessions. Researchers may analyse discourse patterns, reasoning shifts, and qualitative responses across languages and cultures.

The third area, “Governance and Institutions,” focuses on public demand for AI governance structures, regulatory approaches, and institutional trust.

CIP noted that these themes are only illustrative and applicants with strong proposals outside the listed examples may still be considered.

Programme timeline and deliverables

The fellowship will be divided into three phases.

During the first phase, selected fellows will complete onboarding and submit a two-page research plan.

The second phase, running between months two and four, will require researchers to present preliminary findings to the CIP team.

The final phase, covering months five and six, will involve submitting a final output. This may include a published paper, public blog series, or a working prototype.

Compensation will be tied to completion of each milestone.

Who can apply

CIP said it is mainly seeking late-stage PhD candidates and early-career postdoctoral researchers in fields such as:

• Political Science
• Sociology
• Computational Social Science
• Computer Science
• Machine Learning and AI

However, the organisation added that applicants from adjacent disciplines with strong quantitative or mixed-methods experience are also encouraged to apply.

Independent scholars and applicants without institutional affiliation can also apply.

CIP further stated that international applicants are fully eligible for the fellowship.

According to the organisation, selection will focus more on research quality, originality, and motivation rather than institutional prestige.

Application requirements

Applicants are expected to submit:

• A one-page research proposal explaining the proposed research topic, relevant datasets, novelty, and expected impact
• A CV, resume, or LinkedIn profile not exceeding two pages
• A 200–300 word statement outlining experience with large datasets and interest in CIP’s work
• A writing sample such as a paper, working paper, or analytical blog post

Before submitting applications, candidates are advised to explore the publicly available Global Dialogues dataset on GitHub.

CIP said applications will be reviewed on a rolling basis and shortlisted candidates may be contacted within two weeks after the deadline.

The organisation also noted that earlier applicants may be selected before the final deadline.

How to apply

Interested applicants can apply through the official application form provided by CIP.

Application link

About CIP

The Collective Intelligence Project is a nonprofit research organisation focused on building systems and tools for democratic oversight of AI technologies.

Its work includes survey research, deliberative governance methodologies, AI evaluation systems, and preference verification frameworks aimed at improving public participation in AI-related decision-making.