AI Energy Grants

AI chip demonstrators for EU compute infrastructure is sponsored by European Commission — Digital Europe Programme. Expected Outcome: The expected outcomes of this topic are : Functional demonstrators of working AI chip prototypes (PCB-level) designed by EU fabless companies and physically integrated with their system partners. Validation of AI chip solutions on a common EU evaluation platform (provided by the Topic X.B consortium), delivering comparative evidence on key performance indicators. A shortlist of European AI chip solutions awarded the AI Compute Excellence label and qualified to proceed to full rack development under Topic X.A2. Programme areas: DIGITAL

AI compute evaluation and deployment platform for EU infrastructure is sponsored by European Commission — Digital Europe Programme. Expected Outcome: The expected outcomes of this topic are : A common EU evaluation platform for AI chips and racks, with transparent KPIs, workloads and benchmark reports used for Chips JU selection and procurement decisions. First pilot deployments of European AI compute systems validated for EU sovereign partitions of AI infrastructure. Structured collaboration between buyers and suppliers, supporting deployment, integration and optimisation of European AI compute solutions. Stronger EU demand for European AI chips and racks, helping to create a market for deployment-ready solutions. Increased EU technological sovereignty across the AI compute stack for European datacentres. Programme areas: DIGITAL

EQUAL Compute Network is an IDRC-funded initiative to address compute inequities limiting Global South participation in frontier AI research. The network supports researchers, public-interest compute providers, and policy actors in Africa, Latin America, South Asia, and Southeast Asia to evaluate compute needs, build shared compute infrastructure, develop pooled access models, and inform national AI compute strategies. Sub-grants fund research nodes, pilot deployments of shared compute clusters, and policy analysis on compute access. Aligns with the wider AI4D portfolio and complements Canadian and EU compute sovereignty initiatives.

NSF 26-513 establishes State and Regional Artificial Intelligence Infrastructure Hubs to expand researcher access to AI computing, with approximately 100 million US dollars available and about 10 awards anticipated per cycle, one per state or multi-state region, at 4 to 12 million dollars each. The solicitation addresses five core elements: hub consortium governance built on multi-institutional partnerships spanning higher education, industry, philanthropy and state and local government; computing, data, software and other AI infrastructure, funded by the consortia rather than by NSF; regional partnerships including industry collaboration, workforce development and integration with the National AI Research Resource; an AI infrastructure workforce of systems administrators, engineers and AI for science facilitators who support researcher access; and faculty training through instructional materials, labs, workshops and programmes that help students use computing and data resources for scientific research. The diagnosis behind the programme is that the constraint on AI-enabled science at most United States institutions is no longer the existence of compute but the absence of local capacity to make it usable: no facilitators, no trained faculty, no institutional pathway to a national resource. The solicitation is domain-neutral, emphasising broad enablement across science and engineering rather than naming healthcare, climate or agriculture priorities, which means a hub serving an agricultural or biomedical research region can shape its proposal around that regional strength without working against the call. Full proposals are due 4 November 2026, then 3 November 2027 and the first Wednesday in November annually thereafter.

NSF 26-513 establishes State and Regional AI Infrastructure Hubs to widen access to AI compute for scientific discovery, with approximately 100,000,000 dollars available and about 10 awards anticipated per cycle at 4,000,000 to 12,000,000 dollars each over five years. A hub coordinates a state or multi-state consortium and funds three things that are usually unfunded in AI infrastructure efforts: AI infrastructure professionals such as systems administrators, engineers and cybersecurity specialists; AI-for-science facilitators who advise domain researchers on how to actually use AI methods; and faculty training plus instructional material development. It also funds workforce development offering stackable credentials in data engineering, research software engineering and GPU programming. The structural signal in this solicitation is that NSF has concluded the bottleneck in AI-enabled research is human rather than hardware, and is paying for the expertise layer between a researcher and a cluster. The restriction of one award per state or multi-state region makes this a coordination instrument rather than an open competition: institutions within a region are expected to consolidate into a single proposal, which means partnership-building should begin well before the submission window. Deadlines recur on the first Wednesday in November annually, with the current deadline 4 November 2026 followed by 3 November 2027.

National Artificial Intelligence Research Resource (NAIRR) Operations Center (NSF 26-513) is sponsored by National Science Foundation (NSF) with contributions from other federal agencies (e.g., DOE, NIH, NOAA, DARPA, NASA, NIST). The NAIRR Operations Center represents a major funding opportunity, supporting national-scale data infrastructure that powers open, data-intensive and AI-driven research and education.

The NSF State and Regional Artificial Intelligence Infrastructure Hubs program (NSF 26-513) funds state or multi-state regional consortia that combine university, industry, philanthropic, and government partners to widen access to AI computing infrastructure for scientific research. The program's central premise is that the bottleneck in AI-enabled science is no longer algorithms but access - many institutions, particularly community colleges, primarily undergraduate institutions, and smaller universities, cannot reach the compute, data, and expertise that frontier AI research requires. The distinguishing and easily missed feature of this solicitation is what NSF will and will not pay for: NSF funds the coordination layer - consortium governance, workforce development for AI infrastructure professionals such as research software engineers and system administrators, and faculty training - but it does not fund the purchase of computing hardware, data resources, or software licenses. Consortia must demonstrate that they have independently secured those resources from state governments, industry partners, or philanthropy. This makes the program a poor fit for a single institution seeking a GPU cluster and a strong fit for a coalition that already has infrastructure commitments and needs the connective tissue to make them usable across institution types. Awards run $4,000,000 to $12,000,000 over five years, with roughly 10 awards per cycle. Deadlines are 4 November 2026 and 3 November 2027, recurring on the first Wednesday in November annually.

U.S. National Science Foundation State and Regional Artificial Intelligence Infrastructure Hubs is sponsored by U.S. National Science Foundation (NSF) Directorate for Computer and Information Science and Engineering (CISE). This program funds consortia to build advanced AI research infrastructure and workforce training across states and territories, expanding researcher access to AI compute.

NSF 26-513 establishes State and Regional Artificial Intelligence Infrastructure Hubs to expand access to AI compute for scientific discovery, particularly at institutions that have historically been priced out of large-scale GPU resources. The critical structural detail for anyone preparing a proposal is what NSF will and will not pay for: NSF funds the Hub's consortium coordination and governance, regional stakeholder partnerships, AI infrastructure professionals who support the Hub's infrastructure, and faculty training and instructional material development - but it explicitly does not fund the acquisition of the infrastructure itself. Hardware procurement is the responsibility of the regional consortium, which means a credible proposal must arrive with committed non-NSF capital for compute already in hand or clearly pledged. Awards run $4,000,000 to $12,000,000, with approximately 10 awards per cycle against roughly $100,000,000 in total program funding, and only one award will be made per state or multi-state region - so, as with several of NSF's 2026 place-based AI programs, the real competition happens within a state before the proposal is ever submitted. The program pairs compute access with workforce capacity building in AI for science across diverse institution types, including two-year colleges and community colleges, which are eligible as lead proposers. Deadlines recur annually on the first Wednesday in November: 4 November 2026, then 3 November 2027, and annually thereafter. Institutions are limited to one proposal, and individuals to one PI or co-PI role per deadline.

U.S. National Science Foundation State and Regional Artificial Intelligence Infrastructure Hubs: Expanding Access to Compute for Scientific Discovery (NSF 26-513) is sponsored by National Science Foundation (NSF). This program aims to expand access to AI infrastructure for scientific discovery and transform research, education, and training opportunities at institutions of higher education. It supports the creation of statewide or multi-state AI Infrastructure Hubs that coordinate partnerships, expand researcher access to AI computing resources, develop AI infrastructure workforce pathways, strengthen faculty capacity, and create instructional materials. NSF funding supports coordination, workforce development, faculty training, and educational programming, rather than the purchase of AI infrastructure itself. The program also emphasizes opportunities for community colleges, smaller institutions, and regional workforce partners to play meaningful roles in AI ecosystem development.

NSF 26-513, opened 4 August 2026, is a 100 million dollar cooperative-agreement programme establishing State and Regional Artificial Intelligence Infrastructure Hubs. Its purpose is to extend frontier AI compute, data and software capability to institutions and regions that do not currently have it, rather than to deepen capacity where it already exists. NSF anticipates up to ten awards of 4 to 12 million dollars each over five years, and applies a hard geographic rule: only one award per state or multi-state region in this round. That single constraint dominates strategy, because it converts the competition from a national field into an intra-state one - the practical first step for most applicants is to find out who else in their state is preparing a proposal and to consolidate rather than compete. Awards are cooperative agreements, not standard grants, so NSF retains substantial operational involvement in how the hub is run. Proposals must be led by a U.S.-accredited two- or four-year institution of higher education or a U.S.-based non-profit non-academic organisation, and the consortium must include at least one institution of higher education, with private industry, philanthropy and state or local government participation expected. Community colleges are explicitly eligible as lead institutions, which is a deliberate widening relative to most NSF AI programmes. Full proposals are due 4 November 2026, and the solicitation sets the same first-Wednesday-in-November deadline to repeat annually.

The AI Gigafactories Call is the largest single AI funding instrument in Europe and one of the least like a grant. EuroHPC JU is not funding research projects; it is co-investing in the construction of up to seven physical facilities that combine massive computing power with state-of-the-art AI processors, the necessary software and cloud technology stacks, high-bandwidth connectivity and energy-efficient data centres, to provide frontier AI training, fine-tuning and inference as sovereign European services. The applicant profile follows from that: applications are open to consortia or Special Purpose Vehicles bringing together companies, public entities, investors and other partners, and a facility may sit within a single Member State or be distributed across several. This is a call for infrastructure developers and capital syndicates, not for research groups. The financial structure is co-funding rather than full funding - successful projects receive differing levels of EU and national support reflecting the size and range of sovereign services they commit to provide, against an EU and national envelope of up to 10,000,000,000 euros that is expected to mobilise at least 20,000,000,000 euros of private investment. The timeline is long and worth planning around: the submission deadline is 12 November 2026, competitive evaluation follows, selections are expected in early 2027, and operations are to commence within 18 months of selection. Note that the AI Gigafactories are a tier above the existing AI Factories - EuroHPC currently oversees 19 AI Factories and 13 AI Factory Antennas that provide free access to SMEs and startups - so organisations wanting to use European AI compute rather than build it should look to the AI Factories access calls instead. Submission is through the EU Tenders and Funding Portal.

The EuroHPC Joint Undertaking launched its AI Gigafactories call on 30 July 2026 as a joint procurement, not a conventional research grant, and that distinction shapes who should respond. It seeks consortia or Special Purpose Vehicles bringing together companies, public entities, investors and other partners to establish up to seven AI Gigafactories across Europe - facilities that combine massive computing power and state-of-the-art AI processors with the software and cloud technology stacks, high-bandwidth connectivity and energy-efficient data centres needed to deliver frontier AI training, fine-tuning and inference at scale. This is the tier above the 19 AI Factories and 13 AI Factory Antennas EuroHPC already oversees, which serve SMEs and startups with free customised support; the Gigafactories are aimed at frontier-scale sovereign capacity. The financing model is blended: EuroHPC expects total private investment of more than EUR 20,000,000,000 across the EU, with successful projects receiving varying levels of EU and national co-funding calibrated to the size and range of sovereign services offered. Per-facility EU contributions have deliberately not been published, so applicants must build their own capital stack and negotiate co-funding rather than apply against a fixed grant envelope. Facilities may be established in a single Member State or span several, and cross-border developments are explicitly welcomed. The reference is EUROHPC-2026-CEI-AIGF-01 and the deadline is 12 November 2026 at 16:00 CET. Realistically this is a call for infrastructure-grade consortia with credible access to billions in capital, energy siting and hyperscale operations experience - not for individual research groups or early-stage startups, who should instead route through the existing AI Factories network.

The Google for Startups Cloud Program provides Google Cloud and AI compute credits to early-stage companies, with a dedicated AI-first track offering up to 350,000 US dollars in credits to startups building AI-first products, substantially above the up to 200,000 dollars available through the general Scale tier. The programme is a rolling, always-open offering rather than a deadline-driven competition, with applications made through the Google Cloud startups website and decisions returned relatively quickly, which makes it one of the more accessible sources of substantial compute for an AI company without a research affiliation. Eligibility for the AI-first track centres on company age and funding stage rather than on technical merit: the commonly published criteria require the company to have been founded within the last five years and to be at pre-seed or seed stage within five years, or to have raised a Series A within the preceding twelve months, together with an AI-first product. The practical consequence is that this is a window that closes, and an AI company more than a year past its Series A has typically aged out of the largest tier, which argues for applying earlier than a team might otherwise think necessary. Credits are usually accompanied by technical support, training resources and access to Google Cloud's startup technical team. Because terms, tier names and credit amounts are revised periodically and are not always published consistently across Google's pages, applicants should confirm the current offer directly with Google Cloud rather than relying on third-party summaries.

The Enterprise Compute Initiative, announced by Singapore's Ministry of Finance at Budget 2025 with up to 150,000,000 Singapore dollars set aside and administered through Digital Industry Singapore, supports Singapore-based companies undertaking AI transformation projects by providing cloud credits and related tools alongside subsidised consultancy to develop a minimum viable product and manage the organisational change. It is one of the largest national AI compute subsidy schemes outside the research sector, and its design is unusual in pairing compute with expertise: the government co-funds up to 70 percent of eligible consultancy costs capped at 105,000 Singapore dollars per enterprise, paying the consultant's fee directly, while cloud credits of up to 250,000 Singapore dollars per provider, reported as reaching 350,000 for some providers, cover the infrastructure. The eligibility criteria make clear that ECI is aimed at established enterprises ready to build rather than at startups or at companies beginning their AI journey, requiring at least ten Singapore-based staff, an in-house technology, AI or data team of at least two relevant roles, prior experience developing custom AI solutions such as a proof of concept or pilot, accessible and relevant datasets, and CEO-level sponsorship. That last requirement is a deliberate screen: the programme's designers evidently concluded that AI projects without executive sponsorship fail regardless of funding, and they have made sponsorship a condition rather than a recommendation. Applications are made online through the DISG portal and the scheme operates on a rolling basis.

The 2026 ERDC Broad Agency Announcement (W912HZ26S0001) solicits research proposals across a broad range of engineering and scientific disciplines with significant AI and computational focus areas. Issued January 2, 2026, the BAA covers research in AI, computer science, remote sensing, geophysics, telecommunications, hydraulics, dredging, coastal engineering, instrumentation, oceanography, geotechnical engineering, earthquake engineering, vehicle mobility, military engineering, protective structures, infrastructure and environmental issues, energy, facilities maintenance, materials and structures, and ecological processes. ERDC seeks proposals applying machine learning, computer vision, autonomous systems, and AI-driven modeling to advance Army engineering capabilities across both military and civil works applications.

The NSF FDT-BioTech program (NSF 24-561) supports interdisciplinary research at the intersection of AI, computational modeling, and biomedical innovation by funding the mathematical and engineering foundations behind digital twins and synthetic data for healthcare applications. Digital twins — computational replicas of biological systems, patients, or medical devices — require advanced AI and machine learning methods for their development, calibration, and deployment. The program funds research on methods and algorithms relevant to digital twins and synthetic humans, including AI-driven in silico evaluation of medical devices and treatments. Projects must be inherently interdisciplinary, combining expertise in mathematics, engineering, computer science, and biomedical domains. Collaborative projects across multiple organizations are encouraged and can receive up to $1 million in total funding over up to 3 years. The program is administered by multiple NSF directorates including the Division of Mathematical Sciences and the Office of Advanced Cyberinfrastructure, reflecting its cross-cutting nature. The deadline recurs annually on the first Monday in May.

NSF FDT-BioTech Foundations for Digital Twins as Catalyzers of Biomedical Technological Innovation (NSF 24-561) is sponsored by National Science Foundation (NSF). Supports interdisciplinary research at the intersection of AI, computational modeling, and biomedical innovation, focusing on the mathematical and engineering foundations behind digital twins and synthetic data for healthcare applications.

Microsoft Research's global academic research program providing Azure AI compute credits and access to cutting-edge foundation models for research advancing AI safety, human-AI interaction, and scientific discovery. The program operates through periodic Calls for Proposals, workshops, and conferences. Focus areas include AI safety and responsibility (robustness, transparency, evaluation methods), human-AI interaction (trust, creativity, productivity, reducing digital divides), and scientific discovery (knowledge discovery, hypothesis generation, multimodal data generation). The program is evolving into the Agentic AI Research and Innovation (AARI) initiative focusing on intelligent agent systems.

The National Artificial Intelligence Research Resource (NAIRR) Pilot, led by NSF in partnership with DOE, 13 federal agencies, and 28 industry and nonprofit partners, democratizes access to the resources needed for AI research. Rather than cash, awardees receive an integrated ecosystem of compute (HPC allocations and cloud GPU credits), curated datasets, pretrained models, and software platforms. Proposals are reviewed on a rolling monthly basis (submissions by the 15th are typically reviewed by month-end), and projects run for 12 months. A lighter Start-Up track offers a roughly two-week turnaround. All results must be open and publishable. The pilot has supported more than 600 projects and 6,000 students across all 50 states.

The Canada AI Compute Access Fund, administered by Innovation, Science and Economic Development Canada (ISED), provides financial support to help small and medium-sized businesses (SMEs) access the compute power needed to scale and commercialize innovative AI projects. The program has a total budget of $300 million CAD and offers awards from $100,000 to $5 million CAD per project over up to three years. The fund covers two-thirds of eligible costs for Canadian cloud-based AI compute services and half of eligible costs for non-Canadian compute services. Successful applicants receive funding as non-repayable, conditionally repayable, or repayable based on project alignment with public benefits and program goals. The fund is part of Canada's broader Sovereign AI Compute Strategy and complements the AI Sovereign Compute Infrastructure Program. On May 2026, Minister Solomon announced support for 44 Canadian companies across life sciences, healthcare, energy, advanced manufacturing, agriculture, finance, natural resources, and transportation through this fund.

The Scaleway Startup Program provides European early-stage startups with up to €36,000 in cloud credits for AI and machine learning workloads on Scaleway's sovereign European cloud infrastructure. Benefits include access to NVIDIA H100 and L40S GPU instances, Scaleway AI Inference and AI Training managed services, object storage, Kubernetes Kapsule, and serverless compute. The program targets European founders building generative AI, computer vision, and machine learning products who value EU data sovereignty (GDPR-native, hosted in EU data centers) and lower-cost alternatives to US hyperscalers. Participants also receive technical support, architecture reviews, and introductions to the Scaleway and iliad Group startup ecosystem (Station F, Kima Ventures). The program is rolling-application and complements other European compute schemes like OVHcloud Fast Forward.

NAIRR Pilot Start-Up Project Request for Entry-Level AI Compute Access is sponsored by U.S. National Science Foundation (NSF) in collaboration with U.S. federal agencies and industry partners. The NAIRR Pilot Start-Up Project Request is the entry-level track of the National AI Research Resource, giving researchers new to NAIRR quick access to GPU and AI computing resources for proof-of-concept work and scaling studies before submitting a full Research request.

Scaling Compute: AI at 1/1000th the Cost was ARIA's first programme, led by Programme Director Suraj Bramhavar, and it targets a deliberately extreme goal: reducing the hardware cost of training large AI models by more than a thousandfold, while decreasing reliance on leading-edge chip manufacturing. It sits within ARIA's Nature Computes Better opportunity space. The programme was initially backed by 42,000,000 pounds over four years and is now described as backed by nearly 100,000,000 pounds; ARIA has awarded close to 50,000,000 pounds across 12 projects, including a 16,000,000 pound grant to CommonAI for an AI inference lab and work with Oxford engineers on AI system architecture. Two solicitations structured the original intake across four technical areas: TA1 Bold Solutions, TA2 Bold Ideas, TA3 System-level Software Simulation, and TA4 Testing and Evaluation. The programme is explicitly open beyond academia, having funded researchers and engineers across universities, startups, SMEs, large corporates and public labs. The critical status point for anyone assessing this now is that there are no open funding calls for the programme: both prior calls are archived and ARIA directs interested parties to register for opportunity-space updates. It is included here as a live reference for the UK's largest dedicated AI-compute-hardware funding line and because ARIA's model of periodically reopening solicitations within an opportunity space means future calls are plausible. Teams working on analog, photonic, or otherwise unconventional AI accelerators should track ARIA's funding page directly, and compare against ARIA's separate Scaling Trust and Safeguarded AI programmes, which fund different technical areas.

The Biswas Family Foundation Fast Grants Program provides rapid funding for early-stage research projects at the intersection of artificial intelligence and health. Grants of $25,000, $50,000, and $100,000 support 12-month projects focused on biomedical AI, healthcare innovation, computational medicine, diagnostics, scientific discovery, and AI-enabled research tools. The Foundation runs two Fast Grants funding cycles each year, emphasizing speed and support for promising early-stage and high-risk ideas.

ITL Grant Programs (Measurement Science and Engineering (MSE) Research Grant Programs) is sponsored by National Institute of Standards and Technology (NIST). The ITL Grant Program provides financial assistance to support research consistent with ITL's missions, including fields such as Artificial Intelligence, Big Data Analytics, Cybersecurity, and Cloud Computing. NIST's role in AI is to develop the foundational and applied research that leads to trustworthy AI systems, establish technical requirements to cultivate trust in AI systems, and explore future AI computing paradigms.

The Digital Research Alliance of Canada's Accelerated AI Investments program deploys up to $40 million in the 2025-2026 fiscal year to provide dedicated AI compute resources for Canadian researchers. The initiative makes GPU clusters including NVIDIA H100 and A100 hardware available through national HPC facilities including Narval, Cedar, Graham, and Niagara. Resources support AI model training, large language model fine-tuning, computer vision research, and other compute-intensive AI workloads. The program is part of Canada's broader Pan-Canadian AI Strategy which has invested over $2 billion since its launch. Allocations are processed through the Alliance's Resource Allocation Competition (RAC) and Rapid Access Service (RAS) pathways.

The National Artificial Intelligence Research Resource Pilot (NAIRR) is a program from the National Science Foundation (NSF) that funds the creation of an operations center to manage and expand the National AI Research Resource. It provides U.S. researchers and educators with sustained access to advanced AI tools, data, and expertise to support innovation, workforce development, and national competitiveness in artificial intelligence. The program prioritizes research on AI safety, evaluations, and societal impacts, and may provide up to $1,000,000 in compute credits for qualifying AI safety research. Eligible applicants include U.S. researchers, educators, and institutions seeking access to cutting-edge AI computing infrastructure and resources.

The NAIRR Pilot is the single most important entry point for US researchers who need serious AI compute and cannot buy it, and its defining characteristic is that you apply for resources, not funding. Led by NSF with 13 federal agencies and 28 industry and non-profit partners, it was created to democratise access to the compute, data and models that otherwise concentrate frontier AI research inside a handful of well-capitalised labs. The resource catalogue spans GPU clusters (NVIDIA H100 and A100), commercial cloud environments, AI-ready datasets and pre-trained models, drawing on roughly 3.77 exaFLOPS of federal capacity including DOE national laboratory systems. A practical detail worth knowing before writing anything: parts of the catalogue - certain pre-trained models, datasets and platforms - are accessible without a formal proposal at all, so the first step should be checking whether what you need already sits behind an open door. For everything else, the application is deliberately light by federal standards: an electronic form plus a three-page PDF, reviewed by peers against alignment with focus areas, project readiness, technical feasibility, necessity of the requested resources, team expertise and resource justification. The cycle is monthly and continuous - submissions in by the 15th are typically decided by the end of that month - and the call stays open until the pilot programme ends or all resources are committed, whichever comes first. That last clause is the real deadline. The pilot has already supported more than 600 research and education projects reaching 6,000 students across all 50 states, and is transitioning toward a permanent operations centre under NSF solicitation 25-546. One administrative trap catches applicants repeatedly: submissions must come from an institutional email address, and personal accounts such as Gmail are rejected outright.

The NAIRR Pilot Start-Up Project Request is the entry-level track of the National AI Research Resource, giving researchers new to NAIRR quick access to GPU and AI computing resources for proof-of-concept work and scaling studies before submitting a full Research request. Projects receive a three-month allocation of a single resource, with review decisions returned within 2-3 weeks and projects required to begin within two weeks of award notification. Resources include up to 2,000 GPU-hours across systems such as PSC Bridges-2, Purdue Anvil AI, SDSC Expanse AI, and TACC Vista.

National Artificial Intelligence Research Resource (NAIRR) Pilot is sponsored by National Science Foundation (NSF) / U.S. Department of Energy (DOE). The National Artificial Intelligence Research Resource (NAIRR) Pilot is a program from the National Science Foundation and U. S. Department of Energy that provides researchers with access to advanced AI computing infrastructure, software, data, models, and educational resources.

National AI Research Resource (NAIRR) Pilot is sponsored by U.S. National Science Foundation (NSF) in partnership with 13 federal agencies and 28 industry partners. The NAIRR Pilot aims to democratize access to AI compute, datasets, and pre-trained models. Access is open to researchers, educators, and students at U. S. -based academic institutions, nonprofits, federal agencies, tribal agencies, and even startups with federal grants.

Empire AI Consortium is sponsored by New York State, State University of New York System, and philanthropic backers (e.g., Tom Secunda). Empire AI is a New York state project to further artificial intelligence technology research. It supports work on domains such as climate change, drug discovery, education, food insecurity, cybersecurity threats, and healthcare diagnostics. The consortium provides leading public and private research universities with access to state-of-the-art AI computing resources.

National AI Research Resource (NAIRR) is sponsored by National Science Foundation (NSF) in partnership with 13 federal agencies and 28 industry partners. The NAIRR Pilot was created to democratize access to AI compute, datasets, and pre-trained models. It aggregates resources from contributors including Microsoft, NVIDIA, and national laboratories. Nonprofits can gain access to GPU compute resources.

The Google for Startups Cloud Program provides venture-backed, early-stage startups with cloud and AI compute credits to build and scale on Google Cloud. AI-focused startups can receive up to $350,000 in Google Cloud credits (versus $200,000 for other startups) usable over two years, along with access to GPUs and Cloud TPUs, dedicated technical support, training on building generative AI applications with Vertex AI and Gemini models, and Google mentorship. The program is designed to lower the compute cost barrier for startups training and deploying AI models.

NAIRR Pilot (National Artificial Intelligence Research Resource Pilot) is sponsored by U.S. National Science Foundation (NSF) in partnership with 13 federal agencies and 28 industry partners. The NAIRR Pilot was created to democratize access to AI compute, datasets, and pre-trained models. It aggregates resources from contributors including Microsoft and NVIDIA. It is currently transitioning from the pilot phase to a permanent operations center.

U.S.-Qatar Strategic Partnership Initiative: Economic, Technology, and Security Cooperation (Freedom 250 Commemorative) is sponsored by U.S. Department of State, Embassy Doha. This funding opportunity supports projects that strengthen U.S.–Qatar strategic relations and advance U.S. economic and technological leadership, particularly in AI, energy, advanced manufacturing, and critical infrastructure. It seeks to expand opportunities for U.S. exports, investment, and commercial partnerships.

National AI Research Resource (NAIRR) Pilot is sponsored by National Science Foundation (NSF) in partnership with federal agencies and industry. The NAIRR Pilot was created to democratize access to AI compute, datasets, and pre-trained models. It aggregates resources from contributors including Microsoft and NVIDIA. It is currently transitioning to a permanent operations center.

Artificial Intelligence (AI) (NSF SBIR/STTR Topic) is sponsored by National Science Foundation (NSF). This NSF SBIR/STTR topic focuses on cutting-edge technologies in deep learning-based AI systems and AI-based hardware. It emphasizes next-generation AI technologies that are safe, reliable, fair, robust, privacy-preserving, and efficient. This includes AI for societal impact and various sub-topics like cognitive science-based AI, computer vision, conversational AI, and trustworthy AI.

Cassava Technologies and Rockefeller Foundation AI Access for African NGOs is sponsored by The Rockefeller Foundation and Cassava Technologies. This initiative, a collaboration between Cassava Technologies and The Rockefeller Foundation, provides access to AI compute capacity to select Rockefeller Foundation grantees working in various African countries, including Nigeria. The goal is to ensure African-led innovations in sectors like agriculture, healthcare, and education have the resources to improve outcomes with AI. It aims to empower African innovators to build inclusive AI solutions using local datasets, languages, models, and voices.

The EQUAL Compute Network is IDRC's response to the most concrete constraint facing AI researchers in the Global South: not talent or ideas, but access to compute. Launched within the CAD 110 million AI for Development (AI4D) partnership funded by the Government of Canada and the UK's FCDO with additional support from the Gates Foundation, EQUAL works through networks housed in public universities across Africa and beyond, and coordinates with the AI for Development Funders Collaborative to pool resources from G7 governments, multilateral institutions and philanthropies. It funds three distinct things. First, cloud credits - reduced-cost access negotiated on behalf of Global South innovators. Second, hardware and capacity building, covering on-premises compute, technical certifications and data science support, which matters in contexts where bandwidth and foreign-currency billing make cloud impractical. Third, research and networking, including study of what sustainable compute infrastructure actually looks like in low-resource settings and of low-resource innovation methods. Priority goes to AI4D grantees and affiliated labs, and to solutions aligned with the Sustainable Development Goals in health, education and food security. AI4D entered its second phase in 2024 with scope broadened from Africa alone to low- and lower-middle-income countries across South Asia, Southeast Asia and the Pacific. IDRC does not run EQUAL as a single open call with a fixed deadline; access is arranged through AI4D programme channels and periodic calls, so the practical route is to engage the AI4D network and monitor idrc-crdi.ca and ai4d.ai for openings.

NAIRR Operations Center (NAIRR-OC) is a grant from the National Science Foundation (NSF) that funds the establishment and operation of a centralized National Artificial Intelligence Research Resource. The grant supports organizations capable of building and managing the NAIRR-OC infrastructure, which will provide researchers, educators, and institutions across the United States with broad access to AI computing resources, datasets, and tools. Award amounts vary based on scope and are determined through NSF's competitive review process. Eligible applicants are organizations with demonstrated capacity to operate national-scale AI research infrastructure.

Baker Hughes Foundation Grants is sponsored by Baker Hughes Foundation. The Baker Hughes Foundation provides grants to nonprofit organizations globally with giving priorities that include environment and climate, education and opportunity, and health, safety, and wellbeing. While not exclusively focused on AI energy systems, projects within environment and climate categories that align with Baker Hughes' focus on power solutions for industrial and AI energy needs may be considered.

The National Artificial Intelligence Research Resource (NAIRR) Pilot, led by NSF with 13 federal agencies and 28 industry partners, provides U.S. researchers and educators with no-cost access to advanced AI computing resources rather than direct cash awards. Resources include GPU clusters (NVIDIA H100 and A100), cloud environments, AI-ready datasets, and pre-trained models totaling roughly 3.77 exaFLOPS of compute capacity, backed by in-kind contributions such as Microsoft's $20 million in Azure credits. Projects receive 12-month resource allocations on a rolling basis until resources are committed.

NSF's ACCESS (Advanced Cyberinfrastructure Coordination Ecosystem: Services & Support) program provides U.S. researchers, educators, and students with free allocations of high-performance and GPU computing time across a national network of supercomputing resources, in support of AI and machine learning research. Allocations are tiered, with entry-level Explore allocations requiring only a one-page abstract and approved within days, scaling up to larger Discover, Accelerate, and Maximize tiers for compute-intensive AI workloads. ACCESS provides in-kind compute rather than direct cash funding.

EDGE Grant Program is sponsored by State of Delaware. The EDGE Grant program offers funding for STEM and non-STEM businesses in Delaware, focusing on those with growth potential, job creation capacity, and measurable economic impact. It requires dollar-for-dollar matching funds. Small businesses focused on AI, computer vision, public safety, and security threat detection would likely fall under the STEM category.

Computer Science Education Fund Grant is sponsored by Missouri Department of Elementary and Secondary Education. This grant program supports K-12 schools in Missouri to enhance computer science education through grants for curriculum development, teacher training, and resource acquisition. While not exclusively AI, computer science education provides a foundational pathway to AI literacy.

The AI Safety Fund is the grantmaking arm of the Frontier Model Forum, and its notable feature is who funds it: Anthropic, Google, Microsoft and OpenAI as founding frontier-lab members, alongside the Patrick J. McGovern Foundation, the David and Lucile Packard Foundation, Schmidt Sciences and Jaan Tallinn. That composition is the fund's defining characteristic in both directions - it gives grantees proximity to frontier model developers and their threat models, and it means researchers whose work requires independence from those developers should think carefully about the positioning. The fund is capitalised at over 10,000,000 dollars and has run at least two completed grant rounds; a recent cohort of 11 grantees shared more than 5,000,000 dollars, selected from over 100 competitive proposals, implying an acceptance rate around 10 percent. Scope is deliberately narrow and applied. The fund states it supports narrowly-scoped research projects that target urgent bottlenecks, organised into workstreams covering AI-Bio, AI-Cyber, AI-Nuclear and AI Security, with prior RFPs also issued in AI Agent Evaluation and Synthetic Content. This is not general alignment theory funding - proposals that do not map onto a specific, near-term evaluation or mitigation bottleneck in one of those domains are unlikely to fit. The practical difficulty for applicants is that the fund does not publish per-grant amounts, eligibility criteria, or a standing open call with dates; RFPs are issued periodically by topic. Anyone interested should contact info@frontiermodelforum.org or monitor aisfund.org and the Frontier Model Forum updates page rather than waiting for a general solicitation, since the topic-specific RFP windows are the only entry point and they are not always widely advertised.

Canada AI Compute Access Fund for Subsidized Cloud AI Compute for Canadian SMEs is sponsored by Government of Canada – Innovation, Science and Economic Development Canada (ISED). The AI Compute Access Fund, part of Canada's $2 billion Sovereign AI Compute Strategy administered by Innovation, Science and Economic Development Canada (ISED), subsidizes access to cloud-based AI compute for Canadian small and medium-sized enterprises.