AI for Science & Safety Nodes Grant Program

The AI for Science & Safety Nodes grant program is administered by Foresight Institute, a private organization focused on advancing beneficial technologies and long-term human flourishing. Through this initiative, Foresight Institute operates physical hubs in San Francisco and Berlin that provide not only financial grants but also office space, compute resources, and a collaborative research environment. The program is designed to counteract the centralization of artificial intelligence development by fostering a decentralized ecosystem of researchers and builders working on AI-driven science and safety advancements. The primary purpose of this grant program is to support AI-first projects that use artificial intelligence as the central mechanism for progress across a range of scientific and safety domains. These domains include AI for security, private AI systems, decentralized and cooperative AI, AI for scientific discovery and epistemics, neurotechnology and brain-computer interfaces, longevity biotechnology, and molecular nanotechnology. The program emphasizes projects that can accelerate both scientific advancement and safety in parallel, particularly those addressing existential risks associated with AI development. Funding provided through this program typically ranges from 10000 to 100000 dollars per project, with a total annual funding pool of approximately 3000000 dollars. Larger awards may be granted to projects focused on AI safety, while smaller amounts are more common for biotechnology and nanotechnology initiatives. In addition to financial support, grantees may receive access to local compute infrastructure and free office space at one of the program’s hubs. Funding is generally disbursed as a lump sum, although multi-year projects may receive payments in tranches contingent on milestone completion. Overhead costs are allowable up to 10 percent of direct research expenses. Eligibility for the program is broad, encompassing individuals, teams, and organizations, including both nonprofit and for-profit entities. However, for-profit applicants must justify their need for grant funding. The program strongly prioritizes applicants who intend to actively participate in one of the physical hubs, as collaboration and in-person engagement are central to the program’s model. Funding-only applications without participation in the hubs are considered only in exceptional circumstances. Applicants must also be prepared to attend workshops and participate in ongoing community activities. The application process requires submission through an online form, where applicants detail their project, resource needs, and alignment with the program’s focus areas. Applicants are expected to outline their use of AI, specify compute requirements if applicable, and provide a project plan and budget. After submission, proposals undergo an internal review for fit and quality, followed by evaluation by technical advisors. Selected applicants may be asked to provide additional information or participate in interviews. Due diligence checks are required prior to final award decisions. Applications are accepted on a rolling monthly cycle, with deadlines occurring on the last day of each month. The next listed deadline is May 31, and applications are reviewed continuously until capacity is reached. The review process typically takes approximately two months. The program is recurring and ongoing, with new application opportunities each month. Grantees are expected to provide periodic progress updates and comply with reporting requirements related to fund usage and milestone achievement. Evaluation criteria focus on the project’s potential impact on AI safety, feasibility within short development timelines, alignment with program focus areas, and the applicant’s ability to execute. High-risk, high-reward projects are encouraged, particularly those with transformative potential. Preference is given to open source initiatives, although exceptions may be made. For further inquiries, applicants may contact the program via the provided email address.

Emphasize strong alignment with AI safety impact, clear short-term milestones, and demonstrated ability to execute; prioritize AI-first approaches and high-impact outcomes