Who funds them, what they fund and how to apply.
Previously offered grants up to $100,000 over 12 months for people building on the open-source Goose AI agent. The former grants URL now serves general Goose documentation and no current apply path is published. A February 2026 winner post remains on the blog.
Funds AI-first projects that use artificial intelligence as the primary engine for progress in science and safety, combining grant funding with office space and compute at hubs in San Francisco and Berlin. Foresight has paused general rolling applications while it moves to focused RFPs. Until those RFPs are published it is not accepting further submissions.
DARPA and NSF developed AI Forge to fund and manage university-led research in AI interpretability, AI control, and adversarial robustness. Its initial university-capabilities request for information closed June 22, 2026 after receiving nearly 130 responses. No public funding call has opened since that RFI.
Batch-based accelerator for seed-stage AI startups investing $250,000 on an uncapped SAFE plus large compute and tooling credit packages. Four batches totaling 85 funded companies to date; applications for Batch 4 are closed and no later batch is announced.
BARDA's DRIVe division uses EZ-BAA to partner with early-stage companies on health-security innovations. The official EZ-BAA page currently reports no open Areas of Interest and directs applicants to other active DRIVe opportunities.
The Chainlink Community Grant Program funded development teams and researchers building a more functional, accessible, and socially impactful smart contract economy. The documented application route now lives under Chainlink's archives path, so public intake is treated as paused until a current apply page is republished.
12-month paid fellowship placing early-career AI talent with US nonprofits; hosts receive a $10,000 implementation grant plus up to $2,500 in Claude licenses/API credits per fellow. The first cohort closed applications on 17 July 2026; Anthropic has not published a later cohort.
Deep Funding is an AI public-goods mechanism listed by Gitcoin. The 2025 challenge submission window ran February 10 to May 10, 2025. There is no current public apply form as of 2026-08-29.
$12M grantmaking program (funded by Coefficient Giving) for academics and independent researchers on deception, robustness, interpretability, red-teaming, and alignment. FAR.AI is nomination-only pending limited evaluation capacity and has not opened a public RFP.
The Global Innovation Fund is a nonprofit, impact-first investment fund for social innovations improving the lives of underserved and vulnerable people. Its latest published application window closed January 15, 2025; the fund says future calls will be posted after review.
The Mozilla Technology Fund (MTF) gave awards of up to $50,000 plus a 12-month mentorship cohort to open-source projects tackling a themed public-interest technology challenge each round. The latest sourced cohort is the 2024 AI and Environmental Justice round; no later public call is on file.
Small grants to help NumFOCUS's fiscally sponsored and affiliated open-source scientific computing projects improve usability, grow their communities, and speed release cycles. Calls for proposals normally run 3x/year; applications are currently closed while NumFOCUS updates its financial and administrative systems.
Cloud/hardware/software research grants (e.g. up to 30,000 H100 80GB GPU-hours, RTX PRO 6000, or DGX Spark units) for full-time faculty at accredited PhD-granting institutions, across simulation & modeling, AI training, and AI inference/agents/systems software. Currently not accepting new applications.
Principles of Intelligence runs PIBBSS as a three-month interdisciplinary AI safety research fellowship for about 20 researchers. The Winter 2026-2027 Cape Town cohort closed applications on 20 July 2026. The program page is collecting expressions of interest for later cohorts.
Schmidt Sciences' Science of Trustworthy AI program funds technical research to understand, predict, and control risks from advanced AI systems, across three aims: characterizing/forecasting misalignment in frontier models, developing generalizable measurement and intervention, and overseeing superhuman-capability and multi-agent systems. Awardees also receive compute, software engineering support, and frontier-model API credits. Funded in two tiers (up to $1M; $1M-$5M+ over 1-3 years) across sub-tracks including "Understanding Safety in AI Systems" (23 awards) and "Inference-Time Compute" (10 awards).
Status describes whether a program is running. Application methods are recorded separately. Read the funder’s terms before applying.