SPAR Research Program for AI Risks: Remote Fellowship for Aspiring AI Safety and Policy Researchers|Deadline 18 Aug 2026

Artificial intelligence is developing rapidly, creating opportunities across science, business, healthcare, education and government. At the same time, the growing capabilities of advanced AI systems have raised important questions about safety, security, governance and their potential effects on society. For students, researchers and professionals interested in exploring these challenges, the SPAR Research Program for AI Risks is offering an opportunity to gain practical research experience through its Fall 2026 cohort.

The part-time, fully remote research program is designed for aspiring researchers who want to work alongside experienced professionals on projects related to AI safety, policy, security, interpretability, biosecurity and societal impacts.

Application Deadline: August 18, 2026

A Practical Introduction to AI Risk Research

The SPAR program is structured around hands-on research rather than classroom-based learning alone. Participants are matched with experienced mentors and work collaboratively on research projects addressing risks associated with increasingly capable artificial intelligence.

The program can be particularly valuable for individuals who are considering careers in AI safety, AI governance, technical research or related areas but need practical experience to determine which direction best fits their interests.

Participants can dedicate approximately 5 to 40 hours per week, providing flexibility for students, working professionals and others managing existing academic or professional commitments.

The Fall 2026 research period will run from September 14 to December 14, 2026, followed by a Demo Day on December 19.

Research Areas Covered by the Fellowship

SPAR supports research across several disciplines connected to reducing the risks posed by advanced AI.

Participants may explore AI safety, including approaches for ensuring that powerful AI systems behave according to human intentions. Another area is AI policy, which examines how governments and institutions can develop effective frameworks for governing transformative technologies.

The program also supports AI security research, including questions around hardware-enabled governance and protecting AI model weights from theft or misuse.

Other research areas include interpretability, which focuses on understanding how neural networks work internally, as well as biosecurity, where researchers explore ways AI could contribute to biological risks and how those risks might be prevented.

The program also considers the societal impacts of transformative AI, including its potential effects on economies, institutions and communities.

What Participants Can Gain

One of the strongest features of the SPAR program is its emphasis on practical experience.

Fellows work on genuine research questions with guidance from experienced professionals. This can help participants develop research skills while learning how experts approach complex and uncertain problems.

Projects may also result in research papers or preprints, giving participants an opportunity to build evidence of their research capabilities. The experience can be particularly useful for students and early-career professionals seeking to strengthen applications for future research positions, graduate programs, fellowships or roles in AI governance and safety.

Participants will also have access to a community of peers working on related questions. Virtual interactions, research collaboration and networking can help fellows establish relationships with others interested in the emerging AI safety field.

Demo Day and Career Exposure

The fellowship concludes with a Demo Day, where participants can present their work through research posters and presentations.

This provides an opportunity to communicate research findings to an audience connected to the AI safety ecosystem. According to the program information, previous Demo Day activities have included participation from organisations such as METR, Redwood Research, GovAI and MATS.

There may also be opportunities to receive prizes for outstanding posters and presentations. The program notes that $7,000 in total prizes was awarded at the previous Demo Day.

Beyond the formal program, some mentor-mentee relationships continue after the fellowship, potentially leading to further research collaborations, publications and career opportunities.

Who Can Apply?

The fellowship is open to applicants at different career stages, including undergraduate students, graduate and PhD students, and professionals.

Applicants may come from technical backgrounds such as:

• Computer science
• Machine learning
• Mathematics
• Cybersecurity
• Biology
• Physics

The program also welcomes people with policy-related backgrounds, including:

• Law
• International relations
• Public policy
• Political science
• Economics
• Governance

Importantly, previous research experience is not necessarily required. Some projects may have specific technical requirements, but applicants are encouraged to consider opportunities that match their existing knowledge and interests.

Fall 2026 Timeline

The Fall 2026 cohort follows a structured timeline. The mentee application period runs from July 28 through August 18, 2026. Successful mentees are expected to receive decisions on September 4.

The research period begins on September 14 and continues until December 14, with Demo Day scheduled for December 19. Participants may also have the option of continuing their research collaborations after the formal program ends.

How to Apply

The SPAR Research Program for AI Risks offers an opportunity for aspiring researchers to gain practical experience in one of the fastest-growing areas of technology policy and research.

With flexible remote participation, expert mentorship and exposure to real-world AI risk research, the program could be particularly relevant for students, researchers and professionals looking to explore careers in AI safety, governance, security or related fields.

Interested applicants should review the available projects and complete the application before the August 18, 2026 deadline.

Application: Apply through the SPAR AI website

Join Our WhatsApp Community

Disclaimer: Remote Job Opportunities (RJO) is not the  organization offering this opportunity. For any inquiries, please contact the responsible organization directly. Please do not send your applications to RJO, as we are unable to process them. Due to the high volume of  emails we receive daily, we may not be able to respond to all inquiries. Thank you for your understanding.