AI Ethics Grants

CIFAR and the Canadian AI Safety Institute fund Catalyst Project proposals addressing sociotechnical considerations in AI safety. The program supports interdisciplinary research in machine learning applications to science and society, with recent funded projects spanning misinformation combat, trustworthy language models, democratic alignment of AI systems, Indigenous AI governance, and real-world safety in autonomous systems. Designed to catalyze new research areas and collaborations at the intersection of social sciences, humanities, and AI safety.

The IAPS AI Policy Fellowship is a fully funded three-month program for professionals seeking to strengthen practical policy skills and contribute to impactful projects in AI governance and policy. The Summer 2026 cohort runs from June to August 2026 with options to participate in Washington DC or remotely. The program begins with a two-week in-person residency in Washington DC followed by remote or in-person work with weekly mentorship and career development support. Fellows work full-time on independent AI policy projects covering areas such as AI regulation compute governance international AI agreements AI safety policy AI workforce impacts and responsible AI deployment. The fellowship received 240 applications for the 2026 cohort representing a 35 percent increase over 2025. IAPS is a remote-first organization and legally supports fellows in many countries. This fellowship is distinct from the Vista Institute for AI Policy Fellowship which focuses specifically on AI law and from the Cooperative AI Foundation fellowships which focus on multi-agent cooperation problems.

The Mozilla Foundation Democracy x AI Incubator funds technology projects that strengthen democratic institutions and civic participation through responsible AI. This cohort supports 10 projects at $50,000 each for 12 months, with top performers eligible for Tier II funding of $250,000. Projects must address one of three categories: (1) better information systems including verification tools, diverse information sources, and algorithmic transparency; (2) institutional transparency and accountability mechanisms; or (3) civic space protection and expansion including organizing tools, privacy technologies, and surveillance resistance. The incubator provides mentorship, peer learning, and connections to Mozilla's network alongside financial support. Applications require working technology with demonstrated traction, a committed team capable of 12-month execution, and at least partial open-source commitment or a clear roadmap to open source. This is distinct from other Mozilla programs and specifically targets the intersection of AI and democratic resilience.

The Cambridge ERA:AI (Existential Risk and AI) Research Fellowship 2026 is a 10-week immersive research program based at the University of Cambridge designed to support early-career researchers and PhD students exploring frontier AI safety and governance. The program offers a fully funded fellowship with salary, mentorship from leading AI safety researchers at Cambridge, and access to one of the world's premier research environments. Fellows work on original research in AI safety, AI governance, AI alignment, and related areas of existential risk from advanced AI systems. The program commences July 6, 2026 and provides global mentorship connecting fellows with the broader AI safety research community. This is a highly competitive opportunity for researchers who want to make the transition into AI safety and governance research or deepen their existing work in these critical fields.

The Gates Foundation AI to Accelerate Charitable Giving Grand Challenge seeks innovative AI solutions that transform philanthropic giving. The RFP addresses the question: How might AI support donors to give more and give sooner? Projects must address one of three challenge areas: (1) Donor Discovery & Connection — helping donors identify aligned causes through recommendation engines, personalized learning tools, and impact visualization; (2) Convert Intent to Action — reducing barriers preventing motivated donors from completing donations through streamlined giving processes and community-building systems; (3) Foundational Infrastructure — building underlying data systems and standards for philanthropic AI including data pipelines, nonprofit interoperability with AI agents, and fraud detection. Solutions benefiting global health and development in low- and middle-income countries receive priority consideration. Grantees must participate in up to three learning convenings with peer organizations. Applications are submitted through the Gates Foundation portal at submit.gatesfoundation.org.

The Pivotal Research Fellowship is a nine-week AI safety research program (June 29 to August 28, 2026) based at the London Initiative for Safe AI (LISA), with optional extensions of up to six months for strong projects. Fellows receive a GBP 6,000-8,000 stipend, GBP 2,000 housing allowance for non-London residents, London travel coverage, compute resources, and weekday meals. The program offers weekly one-on-one mentorship with established AI safety researchers, dedicated in-person workspace at LISA, research management support, workshops, and speaker sessions. The selection process involves a written application, video interview, mentor-specific work task, and personal interview. Pivotal Research reports that 70 to 90 percent of fellows who applied received extensions in recent cohorts, indicating strong support for continued research development. The fellowship accepts researchers from diverse backgrounds including ML, philosophy, policy, physics, and biology.

AI for Information Security Call for Proposals – Spring 2026 is sponsored by Amazon. Amazon is offering research funding to advance artificial intelligence applications in information security. The program supports research in areas like trustworthy and reliable agentic AI systems, threat detection and response, AI model security, and cloud security.

The Amazon Research Awards AWS Agentic AI Spring 2026 call funds academic research advancing the science and practice of agentic AI systems. The program seeks proposals across four priority themes: no-code and low-code agentic AI solutions for rapid deployment and management, AI-enhanced productivity applications enabling human-AI collaboration, multi-agent systems for secure and effective agent collaboration, and agentic AI for scientific discovery spanning health sciences, chemistry, materials science, and physics. Additional welcome topics include safety and responsible AI for agents, agent customization and post-training optimization, data integration and retrieval systems, enterprise-scale operations and governance, and applications in software engineering, healthcare, finance, media, and consumer assistants. The program supports development of open-source tools and encourages research benefiting the broader AI community. Awards are unrestricted gifts with no intellectual property obligations.

Schmidt Sciences invites proposals for the 2026 Science of Trustworthy AI RFP, funding technical research that advances the science of building trustworthy AI systems. The program addresses three interconnected research aims: understanding why frontier AI systems develop misaligned goals that fail under distribution shift or pressure (Aim 1), creating valid evaluations and interventions to control what AI systems learn (Aim 2), and developing oversight mechanisms for superhuman AI capabilities and managing multi-agent risks (Aim 3). Beyond direct funding, awardees receive computing resources including GPUs and CPUs, software engineering support, API credits with frontier model providers, and access to a research community. The program is open globally and encourages cross-institutional and cross-geographic collaborations. Indirect costs are capped at 10% of total direct costs.

Schmidt Sciences 2026 AI Interpretability RFP is a pilot program seeking new methods for detecting and mitigating deceptive behaviors from AI models. The program focuses on understanding deceptive behaviors from large language models including sycophancy and knowingly giving harmful advice problems that are appearing more frequently in frontier AI systems trained with noisy human feedback. Research areas include developing monitoring and detection methods for model deception creating targeted steering methods for intervening on model truthfulness building visualizations or dashboards that communicate model truthfulness to users applying detection and steering methods to AI debate settings or decision support systems and studying the role of deception mitigations in multi-agent interactions. This program is distinct from the Schmidt Sciences Science of Trustworthy AI RFP which focuses more broadly on understanding and controlling frontier AI risks. The Interpretability RFP specifically targets research on detecting deceptive LLM behaviors and developing practical interventions.

Innovate UK competition funding the creation, curation, annotation, and exploitation of FAIR-data and benchmarks to fuel AI industry growth. Projects must create high-quality benchmarks based on representative dataset slices to enable evaluation of new frontier AI models, while also developing larger AI-ready annotated and curated datasets. Supports collaborative projects that advance the UK AI benchmarking ecosystem and enable responsible AI development through standardized evaluation frameworks.

Defense Sciences Office (DSO) Office-Wide Broad Agency Announcement (HR001125S0013) is sponsored by Defense Advanced Research Projects Agency (DARPA) - Defense Sciences Office (DSO). This BAA solicits proposals for innovative approaches enabling revolutionary advances in science, devices, or systems for national security applications, which can include areas related to AI safety and trustworthiness.

The DeployAI project invites European startups, SMEs, and research organisations to submit mature AI solutions for integration into the AI-on-Demand (AIoD) Platform. The structured three-stage innovation programme supports integration, testing, and real-world deployment of trustworthy European AI technologies. Participants receive financial support, mentorship, infrastructure access, beta-testing opportunities, and visibility through industry events. Aims to accelerate adoption and deployment of responsible AI across European sectors.

Trustworthy Reasoning for AI Cybersecurity Tasks, Operations, and Resilience (TRACTOR) is sponsored by DARPA. The DARPA TRACTOR program funds research on developing AI systems that can perform cybersecurity tasks while maintaining verifiable safety properties, addressing the dual challenge of using AI to enhance cyber defense while ensuring AI-driven security tools do not introduce new …

Making a Difference in Real-World Bioethics Dilemmas Grants is sponsored by The Greenwall Foundation. This program supports research to help resolve important emerging or unanswered bioethics problems in clinical, biomedical, or public health decision-making, policy, or practice. Projects may be empirical, conceptual, or normative and aim for real-world, practical impact. While not exclusively AI-focused, AI ethics can be a relevant area within bioethics.

Coefficient Giving (formerly Open Philanthropy) is a grant from Coefficient Giving that funds evidence-based programs and research in global health, animal welfare, scientific research, and effective altruism causes where philanthropic dollars can achieve high impact per dollar. The fund supports organizations demonstrating rigorous evidence of effectiveness and scalable potential. Eligible applicants include nonprofits, research institutions, and projects aligned with Coefficient Giving's priority cause areas. The fund emphasizes transparency, cost-effectiveness analysis, and funding gaps not addressed by government or traditional philanthropy.

UK ARIA Safeguarded AI Programme for Mathematical and Formal Methods for Quantitative AI Safety Guarantees is sponsored by UK Advanced Research and Invention Agency (ARIA). ARIA's Safeguarded AI programme, led by Programme Director David 'davidad' Dalrymple, funds research combining scientific world models with mathematical proofs to develop quantitative safety guarantees for AI systems comparable to safety assurances in nuclear power and passenger…

U.S. DOT's Small Business Innovation Research Program FY26 Phase I Solicitation is sponsored by U.S. Department of Transportation. The U.S. Department of Transportation (DOT) FY26 Phase I SBIR solicitation is open for small businesses to submit proposals for innovative research and development in transportation, including topics related to AI, safety, and infrastructure. This solicitation covers various operating administrations such as the Federal Highway Administration (FHWA), Federal Railroad Administration (FRA), Federal Transit Administration (FTA), National Highway Traffic Safety Administration (NHTSA), and Pipeline and Hazardous Materials Safety Administration (PHMSA).

The Survival and Flourishing Fund (SFF) supports organizations working on long-term survival and flourishing of sentient life, with a strong focus on AI safety, AI governance, biosecurity, and institutional resilience. Founded by Jaan Tallinn (co-founder of Skype), SFF has distributed approximately $152 million since 2019, with $34.9 million in 2025 alone. The 2026 round is estimated at $20-40 million. SFF uses the S-Process evaluation method involving multiple independent assessors and offers three funding mechanisms: Speculation Grants (rolling basis, smaller amounts for quick-turnaround projects), full S-Process grants (annual competitive round), and Initiative Committee funding. Individual grants range from $10,000 to $4,000,000. Assessors may fund proposals across AI safety, pandemic preparedness, governance, and other existential risk areas.

AI for Good Grant for Ethical and Socially Responsible AI Innovation is sponsored by AI for Good. This award focuses on AI solutions that create social good in areas such as education, healthcare, commerce, environmental protection, and community development. While broad, projects applying AI/ML for dynamic systems in these social impact areas could be relevant.

AI for Good Grant for Ethical and Socially Responsible AI Innovation (Moonshot AI for Good Award) is sponsored by Moonshot. This award is designed for young changemakers (aged 15-30) who are building early-stage AI solutions with real-world impact. It focuses on technologies that can improve society and the environment while promoting responsible and ethical use of artificial intelligence in areas like education, healthcare, commerce, environmental protection, and community development.

AI for Good Grant for Ethical and Socially Responsible AI Innovation (Moonshot AI for Good Award) is sponsored by Moonshot Platform. This award is designed for young changemakers (ages 15-30) who are building early-stage AI solutions with real-world impact, focusing on technologies that can improve society and the environment while promoting responsible and ethical use of artificial intelligence.

AI for Good Grant for Ethical and Socially Responsible AI Innovation is sponsored by Moonshot Platform (through fundsforNGOs). An award for young innovators, engineers, and coders aged 15 to 30 who are building AI solutions that create social good in areas such as education, healthcare, commerce, environmental protection, and community development.

AI for Good Grant for Ethical and Socially Responsible AI Innovation is sponsored by Moonshot Platform. This award supports young innovators leveraging artificial intelligence to create ethical, inclusive, and socially beneficial solutions that address challenges across various sectors. The focus is on practical solutions that solve real problems and generate measurable impact.

Neurotechnology for Safe AI Grant is sponsored by Foresight Institute. This program seeks proposals that leverage neuroscience and neurotechnology to improve AI safety and enhance human capabilities. It emphasizes brain-aligned AI models, secure neurotechnology, and biologically-inspired architectures. The goal is to develop safer, more interpretable AI systems grounded in natural intelligence, while also augmenting human capabilities to stay relevant alongside AI.

NSF's SBIR/STTR Artificial Intelligence topic (America's Seed Fund) provides non-dilutive funding to U.S. small businesses translating cutting-edge AI research into commercial products and services for the public good. The topic emphasizes deep-learning-based AI systems and AI-based hardware that are safe, reliable, fair, robust, privacy-preserving, and efficient. Eight subtopics are supported: cognitive-science-based technologies, computer-vision AI, conversational AI, language-based AI, novel AI hardware, sustainable AI for low-resource and edge environments, technologies for trustworthy AI, and other novel AI technologies. Applicants first submit a Project Pitch; if invited, they submit a full Phase I proposal. Phase I awards up to $305,000 fund feasibility, and successful teams may apply for Phase II awards up to $1.25 million plus matching and supplemental funds.

PROGRAM DESCRIPTION 1. Project Background, Goals, and Objectives Indonesia is the world’s third largest democracy and over half of its population is under the age 30. Because of this large youth demographic, the United States has an opportunity to engage the next generation of Indonesian leaders. For the last several years, however, Indonesian favorability towards the United States has failed to increase despite good government-to-government relations that have resulted in support for trade and military agreements. Meanwhile, as countries such as Russia and China have invested heavily in soft power tactics, their favorability ratings have grown substantially. Adversaries in the region further invest substantial resources to altering public perceptions of America and U.S. government policies through information warfare in both traditional and digital media. The U.S. Embassy Voices for the Future debate team network and championship will support young Indonesian leaders who will help counter growing narratives against the United States that may threaten policy wins. In the age of rapidly developing technology that will fundamentally alter the information space, sharpening the critical thinking, debate, and English language skills of Indonesian youth will equip them to navigate complex information environments, identify anti-American propaganda and media manipulation, and advocate for policies that promote reciprocal trade, regional security, and U.S. - Indonesia cooperation. This program will build upon the popularity of our youth programing, including YSEALI programing, which receives substantially more applications than available slots – allowing more youth with interest in the United States to build positive connections with America. EducationUSA will also hold information sessions on legal U.S. study pathways with students. Project Audience(s): ● Geographic Location: Jakarta, Samarinda, Makassar, Padang. ● Indonesian high school students ages 15-17 with good English proficiency and a demonstrated strong academic record. ● Indonesian high school teachers and debate coaches. Project Goal: Demonstrate U.S. commitment to Indonesia’s next generation of leaders. Strengthen civic literacy among Indonesian high school students through structured debate training and competition. Develop critical thinking, public speaking, and argumentation skills that prepare youth for leadership in civil society, government, and the private sector. Expand English-language proficiency in an applied, communicative context. Build lasting networks among Indonesian youth across regions, fostering national cohesion and cross-cultural understanding. Advance people-to-people ties between Indonesia and the United States by exposing participants to American traditions and values. Project Objectives: ● Objective 1: Establish Voices for the Future debate teams in at least 20 schools through academic programing and coaching focused on critical thinking, public speaking, and argumentation skills that prepare youth for leadership. This objective builds the program's foundation by recruiting a minimum of 20 partner high schools across Padang, Samarinda, and Makassar, plus Jakarta-based high schools through an open call, equipping students with a structured debate curriculum focused on critical thinking, public speaking, and argumentation in English. Each school will have a dedicated coach or faculty advisor to guide students through regular practice sessions using consistent training materials and assessment tools and may form mentorships with a U.S. Embassy Regional English Language Fellow. By embedding debate training directly into school activities, the program ensures that young Indonesians across multiple regions develop the civic and leadership skills they need to contribute meaningfully to their communities and country. ● Objective 2: Strengthen teacher and coach capacity through Training of Trainers (ToT) This objective ensures the program's long-term sustainability by building the skills of the educators. Selected teachers and coaches will participate in a Training of Trainers program that equips them with practical tools in debate coaching, facilitation, and civic education. Participants will also form a peer network to share resources and support one another across regions. By empowering educators, the program creates a lasting foundation that can grow and thrive well beyond the project period. ● Objective 3: Increase student engagement on free speech, digital freedom and AI Through debate topics, workshops, and guest speaker sessions, students will explore themes such as free expression, digital rights, misinformation, and artificial intelligence. They will be encouraged to develop their own initiatives to share what they have learned with wider audiences. This prepares Indonesian youth to think critically about the digital world and engage confidently with the democratic values the United States promotes globally. ● Objective 4: Hold a Voices for the Future National Debate Championship in Jakarta and generate public awareness This objective brings the program to a national stage by hosting a championship in Jakarta that unites top teams from all participating schools. A merit-based qualification process will ensure fair competition, with judges drawn from the U.S. Embassy, Indonesian government, civil society, and the private sector. A strong public awareness campaign through social media and press outreach will amplify the event's visibility and celebrate student achievement. The championship will serve as a powerful, public demonstration of U.S. investment in Indonesia's next generation of leaders. Program Activities: · Program design and partnership development The recipient will partner with at least 20 schools across Jakarta, Samarinda, Padang, and Makassar and local debate communities. In coordination with the U.S. Embassy, the recipient will brief U.S. experts on the Indonesian educational context, co-develop coaching curricula and debate topic frameworks, produce a program operations manual, and develop marketing and outreach materials to recruit student and teacher participants. · Debate bootcamp and Coach Training (ToT) Using funding from this grant, U.S. experts will travel to Padang, Samarinda, and Makassar to deliver three training of trainers (ToT) workshops and debate bootcamps. The ToT component will train a minimum of 24 teachers and coaches in debate formats and rules, argument construction and rebuttal techniques, and subject matter knowledge on AI governance, digital freedom, and free speech rights, equipping coaches to independently sustain competitive debate programs beyond the grant period. Each bootcamp will also include a demonstration debate session with recruited students. Coaches will receive a comprehensive toolkit of lesson plans, research guides, practice drill frameworks, and assessment rubrics. EducationUSA will hold information sessions on U.S. study pathways at each bootcamp, with a minimum of 120 students engaged across all three cities. · National Debate Championship in Jakarta A national championship will be held in Jakarta open to two tracks of participants: top-qualifying teams from the Padang, Samarinda, and Makassar bootcamp cities, and Jakarta-based high school teams registered through the open call. All teams will compete across multiple elimination rounds judged by U.S. experts, Indonesian academic and civil society judges, and U.S. Embassy representatives. The program will culminate in an awards ceremony with trophies, certificates, and formal commendations, followed by a closing U.S. expert mentorship session connecting debate skills to democratic citizenship. Top performers from both tracks will be connected with U.S. educational opportunities and EducationUSA advising. A comprehensive media strategy will generate at least 10 placements across social media, press, and digital channels. 2. Substantial Involvement The U.S. Embassy will serve as the Federal awarding agency representative and will maintain substantial involvement in the implementation of this cooperative agreement, consistent with the nature of a cooperative agreement as distinct from a grant. Specifically, the Embassy will: 1. Program Oversight & Coordination: Provide overall programmatic oversight and serve as the primary point of contact for the recipient throughout the period of performance. The Embassy’s Public Affairs Section (PAS) will designate a Grants Officer's Representative (GOR) to monitor program activities and ensure alignment with U.S. foreign policy objectives and public diplomacy goals. 2. Strategic Guidance: Collaborate with the recipient to ensure program activities reflect U.S. Embassy priorities, including the promotion of critical thinking, civic engagement, English language skills, and people-to-people ties between the United States and Indonesia. 3. Approvals & Modifications: Review and approve key program deliverables, including but not limited to: the final program design, participant selection criteria, curriculum and training materials, event logistics for the national championship, and any proposed modifications to the scope of work or budget. 4. Stakeholder Engagement: Facilitate introductions and coordination with relevant host-country government ministries (e.g., Ministry of Primary and Secondary Education), local educational institutions, and other key stakeholders as appropriate to support program implementation. 5. Monitoring & Evaluation: Conduct site visits, review progress and financial reports, and assess program outcomes against established performance indicators. The Embassy reserves the right to request additional information or documentation as needed. 6. Communications & Visibility: Advise the recipient on U.S. government branding and communications requirements. The Embassy may participate in or co-host program events, including the national championship, and may publicize program activities through Embassy communication channels. 7. Financial Oversight: The Grants Officer (GO) will retain authority over all financial matters, including the review and approval of budget modifications, the authorization of payments, and the closeout of the award. participants. Funding Opportunity Number: PAS-JAKARTA-FY26-06. Assistance Listing: 19.040. Funding Instrument: CA. Category: ED. Award Amount: $50K – $150K per award.

Multi-agent AI safety research funding call is sponsored by Google DeepMind, Schmidt Sciences, Cooperative AI Foundation, Advanced Research and Invention Agency, and Google.org. This funding call focuses on technical research into how large-scale multi-agent AI systems behave as a group, and how to provide frameworks to understand and mitigate potential risks.

Investing in Multi-Agent AI Safety Research is sponsored by Google DeepMind, Schmidt Sciences, Cooperative AI Foundation, Advanced Research and Invention Agency, and Google.org. This funding call supports technical research on strengthening multi-agent safety, focusing on how large-scale multi-agent AI systems behave as a group and how to understand and mitigate potential risks. The goal is to build a safe foundation for a multi-agent future.

Schmidt Sciences Scaling AI Safety for a Multi-Agent World 2026 RFP funds technical research advancing the safety, oversight, and control of large-scale multi-actor AI agent ecosystems. The RFP focuses on emergent risks from interacting AI agents including collusion, deceptive coordination, principal-agent failures, multi-agent reinforcement learning safety, scalable oversight for agent swarms, and benchmark development for multi-agent safety evaluation. Tier 1 funds focused investigations up to $300K; Tier 2 funds broader programmes up to $1M over 1-2 years. Strong fit for academic AI safety labs, applied research nonprofits, and frontier safety teams. Deadline August 8, 2026, with applications via Schmidt Sciences online portal.