We're a part-time, remote research fellowship that enables aspiring AI safety and policy researchers to work on impactful research projects with professionals in the field.
Meet our 120+ mentors for Spring 2026
A part-time, remote research program pairing aspiring researchers with professionals addressing risks from AI. Mentees gain research experience and guidance; mentors get capable collaborators.
Commit 5–40 hours/week, depending on your availability. SPAR is designed to fit well alongside other commitments.
3 months of structured research, culminating in Demo Day — posters, talks, and a career fair with organizations like METR, Redwood Research, GovAI and MATS.
We accept mentees with relevant technical or policy background at any level, from undergraduate students to mid-career professionals.
SPAR research has been accepted at ICML and NeurIPS, covered by TIME, and led to full-time job offers for mentees.
Check out some of our work:
SPAR supports a broad range of research that we think is helpful for helping society mitigate the risks posed by advanced AI, including but not limited to:
Ensuring powerful AI systems follow human intentions
Governing the development of transformative AI
Hardware-enabled governance and preventing the theft of model weights
Reverse-engineering neural networks and their internal representations
Preventing or preparing for AI-enabled bioattacks
Understanding the effect of transformative AI on society and the economy
Build the skills to break into technical AI safety, AI policy or biosecurity research and learn whether this path is right for you. SPAR and your mentor provide structured support to help you start your journey.
Get advice and support from experienced researchers to support your development of key scientific skills.
Most projects in SPAR result in papers. Our demo day gives you the opportunity to present your research in front of researchers from organizations like METR, Redwood Research, GovAI and MATS.
Join a research team of peers with shared interests who support each other's growth. Virtual coffee chats and in-person meetups provide additional opportunities to connect.
Conduct real, cutting-edge research to help the world answer the most pressing questions of our time. Most projects lead to published papers or preprints with mentee co-authors.
Last year we awarded $7,000 total at our Demo Day, which rewards the best posters and talks for projects in the program.
From application to publication, here's what you can expect.
STEP 01
Explore available projects and apply to work on ones you're excited about.
STEP 02
Spend 5-40 hours/week on real catastrophic risk research with your team. Receive structured support and regular feedback.
STEP 03
Present your research at our virtual mini-conference attended by leading AI safety organizations. The best posters and lightning talks will be eligible for prizes!
STEP 04
Many mentorship pairs continue working together after the program ends, leading to publications and career opportunities.
I think SPAR is the best place for anyone to get involved and just get started with research.
Kushal, Information Safety
[My mentor] was extremely willing to share his connections and was involved in the trenches with us from the beginning. He provided some thought-provoking insights and guidance and opened himself up for 1:1s, even as he started a new key role... It was a pleasure to work with [my mentor] and I hope it won't be the last time.
Noah, Overview of Systemic Economic Risks from TAI Systems
Thank you for such a fantastic opportunity! During the program I was offered an AI Safety role at Google and recently started right as SPAR ended. I felt like the program gave me great experience and helped strengthen my passion which no doubt helped with getting the job.
Greta, Near Zero-knowledge Detection of Undesired Behavior
It's exactly the "foot in the door" one needs when transitioning into research positions.
Harry, Using RL to train shutdownable agents
The program is open to undergraduate, graduate/PhD students, and professionals of different experience levels. Mentors often look for a technical background (e.g., ML, CS, cybersecurity, math, biology, physics, etc.) or knowledge relevant to policy and governance (e.g., law, international relations, public policy, political science, economics, etc.).
Some projects may require additional background knowledge, like knowledge of specific research areas, techniques, or tools (e.g. ML frameworks).
We don't require previous research experience, and many mentees have none.
Even if you do not entirely meet these criteria, we encourage you to apply! Many past mentees have been accepted even though they didn't completely match a project's criteria.
The SPAR program follows this timeline:
