Apart Research is a remote-first AI safety research organization founded in 2022 that converts technical talent into published AI safety researchers. The organization operates a three-tiered pipeline: weekend research sprints (hackathons) that engage hundreds of participants globally, a Studio program offering 4-6 weeks of structured research development, and a Fellowship providing 4-6 months of intensive research support leading to peer-reviewed publication. Apart has engaged over 5,000 participants across 45+ hackathons in 50+ global locations and produced 22 peer-reviewed publications at venues including ICLR, NeurIPS, ICML, and ACL.
Apart Research is a remote-first AI safety research organization founded in 2022 that converts technical talent into published AI safety researchers. The organization operates a three-tiered pipeline: weekend research sprints (hackathons) that engage hundreds of participants globally, a Studio program offering 4-6 weeks of structured research development, and a Fellowship providing 4-6 months of intensive research support leading to peer-reviewed publication. Apart has engaged over 5,000 participants across 45+ hackathons in 50+ global locations and produced 22 peer-reviewed publications at venues including ICLR, NeurIPS, ICML, and ACL.
Funding Details
- Annual Budget
- $954,800
- Monthly Burn Rate
- $79,567
- Current Runway
- -
- Funding Goal
- $954,800
- Funding Raised to Date
- -
- Fiscal Sponsor
- Ashgro Inc
Theory of Change
Apart Research believes that the primary bottleneck in AI safety is not a lack of problems to solve but a lack of talented researchers working on them. Their theory of change centers on building a scalable pipeline that identifies, trains, and places technical talent into AI safety careers. By running low-barrier-to-entry weekend hackathons globally, they cast a wide net to discover researchers with aptitude and motivation. Those who demonstrate fit progress through increasingly intensive programs (Studio, then Fellowship) that provide mentorship, infrastructure, and support to produce peer-reviewed research. This creates a self-reinforcing cycle: published research contributes directly to the field's knowledge base, while researchers who complete the pipeline go on to join AI safety organizations, expanding the field's capacity. The remote-first, distributed model allows them to reach talent globally rather than only in traditional AI hubs, and the decreasing marginal costs of their infrastructure allow them to scale impact efficiently.
Grants Received
from Open Philanthropy
from Survival and Flourishing Fund
from Open Philanthropy
Projects
AI Safety Ideas is a collaborative, open-access platform for sharing and discovering AI safety research project ideas, built and maintained by Apart Research.
Apart Sprints is the research hackathon program of Apart Research, running weekend-long AI safety research events with thousands of global participants. Teams work on frontier AI safety questions and produce open-source research, evaluations, and prototypes.
A weekly newsletter, podcast, and YouTube series by Apart Research covering the latest research in machine learning and AI safety. The series was later discontinued as Apart Research reprioritized its activities.
People– no linked people
Discussion
Sign in to join the discussion.
Key risk: The main risk is that a hackathon-driven discovery model optimizes for speed and publishability rather than deep alignment progress, creating low counterfactual impact or duplication with programs like MATS and risking quality dilution as a small team scales mentorship and research standards.
Details
- Last Updated
- Apr 2, 2026, 10:09 PM UTC
- Created
- Mar 18, 2026, 11:18 PM UTC
Case for funding: Fund Apart Research because their globally distributed hackathon-to-fellowship pipeline has already converted experienced engineers into safety researchers, yielding 22 peer-reviewed papers (including an ICLR Oral for DarkBench) and 30 placements across top labs and AISI, making it a cost-effective way to rapidly scale technical AI safety capacity beyond traditional hubs.