Currently 50 active paid ai and ai safety grants, fellowships and residencies. Hand-curated and updated weekly. Almost every entry is funded; a few notable unpaid open calls and festival submissions are included as clearly flagged exceptions. Browse the list below, or use the interactive desk for filtering and shortlisting.
Subscribe: RSS feed · Calendar (.ics)
The third round of Project Tech4Dev's AI Cohort, a hands-on implementation programme for NGOs designing, building and deploying responsible AI solutions inside their own programmes. Cohort 2.0 worked with seven NGOs across education, civic engagement, youth development and nonprofit technology, who built and piloted AI teaching assistants, volunteer coaches, mentoring tools and decision-support systems; the organisers' stated lesson was that adoption turns on understanding user needs, testing rapidly, measuring impact and building responsibly rather than on the technology itself. Cohort 3.0 is aimed at organisations ready to move past exploration into solutions that can be deployed in real-world settings. STRUCTURE (three sprints over roughly six months): Discovery Sprint (months 1 to 2) covering problem discovery, use-case prioritisation, data-readiness assessment, user research and solution design; Build Sprint (months 2 to 5) covering prototype development, implementation sprints, user testing, mentorship and responsible AI evaluations; and Activation Sprint (month 6) covering pilot deployment, impact assessment, operationalisation planning and a final showcase with peers and ecosystem partners. SELECTION CRITERIA: NGOs must show some in-house technical or data capability, ideally a dedicated person who can anchor the work (not necessarily an AI expert, but able to manage workflows, communicate operational requirements and work with mentors and experts); must commit one person's full bandwidth for the whole programme, including aligning leadership, adapting operations and making room to test, monitor and refine; and preference goes to problems involving multi-modal input, especially Indic languages, and to use cases built on LLMs. Applications are open to nonprofits across sectors, with priority to climate-focused organisations whose AI use cases could improve programme delivery, decision-making or community outcomes. Apply via the Google Form linked from the announcement before 10 August 2026.
GovAI is hiring for two research roles with a single shared application (apply once to be considered for both). RESEARCH FELLOWS play a central role in GovAI's research: experienced researchers who conduct independent, high-quality work to inform high-stakes decisions in government and industry, whether through policy memos, blog posts, academic publications or strategic advising, alongside mentoring early-career researchers (such as Summer and Winter Fellows) and helping shape GovAI's research agenda. Two-year contract with the possibility of renewal; Fellows work under a Senior Research Fellow or leadership-team member but have substantial autonomy over projects, agendas and publication formats. RESEARCH SCHOLARS join for a one-year visiting position designed to support the career development of AI governance researchers and practitioners: conducting policy, social science or technical research, advising policymakers, or starting new applied projects. TOPICS include risk management, threat modeling, economics of AI, geopolitics of AI, public policy and technical governance; candidates working on adjacent topics are also welcome. LOCATION: London or Washington DC preferred, but open to candidates based elsewhere (for example the Bay Area). ELIGIBILITY: open to a wide range of academic and professional backgrounds (computer science, political science, public policy, economics, philosophy, law, plus government/industry/civil-society experience); no strict educational requirement, but the strongest Fellow candidates typically have multiple years of relevant research or policy experience and a track record of high-quality, impactful work. PROCESS: written application; a paid remote work test (some workstreams add a research agenda and a second paid test); then a final stage of two interviews and reference checks. Deep expertise in technical governance, frontier safety design, cybersecurity or US policy is especially welcome. Deadline 23:59 Anywhere on Earth, Sunday 16 August 2026. Contact recruitment@governance.ai with location or access questions.
A postdoctoral position in the Human Media Interaction group at the University of Twente on the project 'Balancing Personalisation and Polyvocality: Dialogic Interaction for Open-mindedness and Empathy in Conversational AI'. The work develops and evaluates multimodal conversational agents for museum settings that hold personalisation in tension with exposure to plural perspectives, rather than optimising for engagement alone. Tasks include building models that use gaze, speech and dialogue data to infer visitor engagement and openness, designing adaptive dialogue strategies, running VR-based user studies, and publishing in strong venues. Of the current Dutch academic postings this is the closest thing to a funded creative-technologist brief: museum interpretation, multimodal sensing and dialogue design in one role. REQUIREMENTS: a PhD in computer science, AI, HCI or a related field; experience with conversational AI, dialogue systems or adaptive systems; experience running empirical research, preferably user studies; strong analytical and academic writing skills; and excellent English. Apply with a CV including publications, a motivation letter and two academic references. Contact: Shenghui Wang (shenghui.wang@utwente.nl). NOTE: the AcademicTransfer listing page shows a 16 August 2026 deadline while the detail page states 17 August 2026, so treat 16 August as the effective cut-off.
Annual essay competition from the Berggruen Institute, a companion to the Berggruen Prize for Philosophy & Culture, seeking new thinking and cross-cultural perspectives on fundamental philosophical questions of significance for contemporary life and the future. It looks for essays in the tradition of Rousseau, Montaigne, and Emerson (models cited include Frankfurt, Nagel, Sontag, Fukuyama, and recent Noema pieces), not peer-reviewed academic work: novel ideas, clearly argued in compelling ways for intellectually serious readers. THEME for 2026: 'A New Axial Age?' Referencing Karl Jaspers's Axial Age (c. 800-200 BCE), the prompt asks whether civilisation is again at a 'hinge of history' undergoing a pivotal transformation of consciousness driven by rapid technological change and its social, economic, and political consequences. Suggested (non-exhaustive) angles: the foundational ideas and institutions underpinning present and future societies; the philosophical implications of the current era; the future of human evolution and life amid the ecological crisis; shared conceptual foundations for a common future; global transformation driven by technology and the rise of AI; the end of the global order and its ramifications; and the future of life amid biotechnology and synthetic life. Entrants may choose their own topic; for 2026 creative fiction is accepted. ELIGIBILITY: open to the public; not for academic papers and not governed by academic disciplinary standards or affiliations. Length up to 10,000 words in English or 15,000 characters in Chinese (excluding diagrams, tables, endnotes, bibliography, or authorship declaration). One submission per year per entrant, in either English or Chinese; co-authorship allowed. Submissions must be original and previously unpublished. AI USAGE: generative AI must not be used to create content or substitute original thought, and entrants must engage primary sources rather than AI summaries; AI is permitted only for proofreading/grammar, as quoted supporting material when AI itself is the subject of the research, or for translating an essay into English or Chinese, and all AI use (including AI-assisted translation) must be disclosed on the front page or the entry is disqualified. JUDGING: originality, clarity, and style. Essays are due by 23:59 PT on 17 August 2026; PDF only. Chinese-language submissions use a separate page. Inquiries to essay@berggruen.org.
An eight-week full-time empirical AI alignment research programme run by the AI Alignment Foundation, with a dedicated research manager, mentorship from senior researchers, compute and infrastructure, culminating in a Demo Day where fellows present their work. Fellows receive a USD 12,000 stipend paid in two installments plus a compute budget. The programme is fully remote for the research period; Demo Day is hybrid in Los Angeles and travel is covered for fellows who attend in person. ELIGIBILITY: open to all career stages, including recent PhDs, graduate students, self-taught researchers and career-changers, but candidates need a strong machine learning background, demonstrated engagement with alignment research, prior independent research output, and the ability to work autonomously. Non-US applicants are welcome provided they can make Pacific Time meetings; current graduate students may apply if they can confirm availability. This is explicitly a technical programme: the foundation says policy-focused applicants should not apply. SELECTION: all applications are reviewed, qualified candidates advance to a technical test, and finalists are interviewed, with decisions expected by early September. Applications close 17 August 2026 (Anywhere on Earth); the programme starts 8 September 2026.
Antikythera, a research programme on planetary computation and the evolution of intelligence, invites contributions to Agentworld, a special issue of its peer-reviewed MIT Press journal launching Fall 2026. THEME: Agentworld is a research initiative on the medium-term future of hybrid human-AI societies, framed not around tools for agentic task completion but around what the brief calls 'the human perplexity of deep agentic diffusion'. It treats human societies as the evolutionary niches into which new artificial species must adapt, where the constraints of agentic integration are qualitative and open-ended (psychological, philosophical, linguistic, symbolic, existential) rather than utility functions to be solved. The framing describes a 'parasociety' of densely reticulated social graphs in which agent-to-agent communication constitutes most socially significant interaction, humans and agents anthropomorphise each other by inheritance rather than necessity, and the two overlapping societies act as mutual evolutionary niches. The disciplinary posture proposed is 'a preemptive anthropology of open-world centaur societies': building the vocabulary, scenarios and conceptual armature for an emergent sociality while it is still forming, rather than describing it after the fact. WHAT THEY WANT: theoretical essays and technical papers that engage the Agentworld themes directly and explain how the contribution adds to the special issue's core questions. Philosophical, theoretical, technical, design-led and product-oriented research are all in scope. The call explicitly favours work that exceeds conventional academic journal formats, including contributions combining philosophical argument, technical analysis, design research, diagrams, software, infrastructure proposals, visual media or web-native formats. Contributors may optionally propose still images, diagrams, video, web-based or interactive components, and should say whether they have the design, development or production capacity to realise them; these elements are not required. SUBMIT: an abstract or proposal of up to 500 words; author name(s), affiliation(s), contact details, biography(ies) and CV(s); a designated lead author or contact person for collaborative submissions; and a portfolio with at least one relevant work sample such as a preprint, published article or paper. Full paper drafts are encouraged where available but not required at the initial stage. Submissions may be individual or collaborative. TIMING: the review window opened 25 June 2026. NOTE: no closing date is published on the call page; the deadline shown here is a placeholder set roughly a month out, so check the site or submit early.
A $5,000 fellowship to spend six weeks turning a very large Canadian public archive into something new. THE MATERIAL: Internet Archive Canada's Government Publications collection, part of the Democracy's Library project, holding over 104,000 text items, roughly 22.1 million scanned pages and about 48.25 TB of Canadian public-record material, browsable at archive.org/details/governmentpublications. WHAT TO BUILD: deliberately open-ended. The call names a search tool that answers a real question with page-level citations, a timeline tracing one policy across decades of government language, and a sonic or visual piece built from archival material, but explicitly invites work nobody has thought of yet. The requirements are that it is useful, that it is imaginative, and that the code is open source when it is done. ORIGIN: the fellowship came out of conversations between BC + AI's Kris Kruger and IAC executive director Andrea Mills about treating a public archive as material for new work rather than only something to preserve. IAC brings the archive, BC + AI brings a community of builders used to working in public. ELIGIBILITY: anyone, anywhere in the world; one builder is selected. APPLICATION: a specific project description grounded in the collection, a phase-by-phase work plan for the six weeks, and a success metric, submitted at bc-ai.ca/fellowship/apply. KEY DATES: applications close 21 August 2026 at 11:59pm Pacific; fellow selected by 31 August; build runs 1 September to mid-October; showcase at Futureproof Festival in Vancouver on 28 October 2026 if the project is ready.
Collider, the New York coworking and convening space for AI safety and other high-impact work, is hiring a Program Manager to design, run and scale the programmes that grow the AI safety talent pipeline and infrastructure in New York. The role owns programmes end to end, from concept to launch to iteration, working with researchers, practitioners, educators, funders and partner organisations. WHAT YOU DO: design and operate fieldbuilding programmes (fellowships, training programmes, workshops, conferences, cohort courses and new formats); own budget, timeline, vendor relationships, participant experience and post-programme evaluation; recruit and coordinate mentors, instructors, advisors and participants, and build the pipelines and selection processes that hold quality as things scale; define and track outcomes with metrics beyond participant satisfaction, and iterate on the evidence; maintain relationships across research labs, university groups, training programmes (MATS, ARENA, BlueDot, SPAR, AI Safety Camp are named), grantmakers and adjacent communities; represent Collider externally; write programme reports, strategy memos, post-mortems and public materials; and work with leadership on which programmes to launch, scale or sunset. REQUIRED: 3+ years of programme or project management, ideally running cohort, fellowship or training programmes; substantive understanding of the AI safety field, its organisations, its main research agendas (alignment, interpretability, evaluations, governance) and its open problems; strong operational instincts; an entrepreneurial, initiative-taking approach to judgement calls about people, partnerships and priorities; and clear written communication. NICE TO HAVE: prior experience at an AI lab, AI safety organisation, university group, training programme or grantmaker; a background in education, talent development or community building; enough ML or alignment literacy to follow a paper's argument; policy literacy on the AI governance landscape; or a background in finance, law or another professional field Collider wants to build pipelines into. PROCESS: apply via the form in the job document by 23 August 2026 (applications are reviewed on a rolling basis and an offer may be made before the deadline, so applying early is advised), then a 15 to 25 minute screening call with hiring partner Impact Ops, an interview with the prospective manager, a paid one-day in-person work trial at the Collider coworking space, and 2 to 3 references. Reports to the Director.
A postdoctoral position in the Philosophy section at the University of Twente, part of the 4TU.Center for Ethics and Technology, working on the ethics and epistemology strand (Work Package 6) of LLMs4EU. LLMs4EU is a European Commission funded initiative led by the Alliance for Language Technologies, spanning more than 20 countries, to build large language models across all EU languages and preserve European linguistic and cultural diversity in the digital age while keeping the resulting tools compliant with European values and legislation including the AI Act and the GDPR. TASKS: establish benchmarking frameworks that address legal and ethical concerns; develop guidelines for fine-tuning processes with an emphasis on preserving cultural heritage; oversee ethics compliance and advise the project ethics board; and build frameworks that integrate ethical considerations into LLM design. This is the tech-communication and language-politics end of AI research rather than a modelling job: the questions are whose language, whose culture and whose norms get encoded when a continent builds its own models. REQUIREMENTS: a doctorate in ethics, philosophy, science and technology studies, AI ethics, computer science with an ethics focus, or a related field; demonstrated expertise in AI ethics and ethics compliance; understanding of AI tool design and LLM development; appreciation for cultural diversity; strong communication and organisational skills; and professional English. Apply through the AcademicTransfer posting. Closes 25 August 2026.
The annual DARC PhD/Research Workshop, organised by the Digital Aesthetics Research Center at Aarhus University with transmediale, responds to transmediale's 2027 thematic framework. THEME: 'Interstices: Research (in) the gaps'. An interstice is the small space between objects set closely together, the cords in a rope or the atoms in a crystal lattice: an emptiness produced by the structure it belongs to, and without which that structure could not be what it is. In crystallography, interstitial sites are the spaces a lattice and its basis generate but do not fill, and interstitial diffusion, the movement of small atoms through the gaps of a host lattice, is the mechanism by which materials transform. The call asks what the equivalent sites are in social, cultural and computational structures: where are the gaps these combinations generate but do not fill, what moves through them, and do they carry transformative power? It works through Yancey Strickler's dark forest theory of the internet (the withdrawal of meaningful exchange from the open, surveilled, indexed web into private channels), the Dark Forest Collective's anthology and its 'cozyweb' of small sheltered spaces, dark nets, miniverses and moving castles as interstitial spaces produced by and defined against the dominant web, and Bogna Konior's extension of the dark forest metaphor to the opacity of AI systems, where silence is refusal rather than absence and the forest is increasingly inhabited by other-than-human intelligences that opt out of engagement with people. Where big tech pursues efficiency through the automation of human perception, cognition, identification and aesthetics, artists and researchers explore the non/more-than-human, glitchy, noisy, machinic, datafied and dark to build new expression and critique of data-ism and what is marketed as artificial intelligence, including the dangers and potentials inside badly curated data models, between polluting data centres, and in skeuomorphic interfaces. WHAT THEY WANT: proposals from artists, researchers and practitioners for work that occupies and moves through gaps that are social as well as structural, ontological and material. The gaps are physical too: the heat of data centres and their thirst for water, latency in the cables, unindexed nodes. The call invites speculations, practical interventions and inferences that fill the gaps and expand the diversity of possible communications and communities that can inhabit dark forests, including those built on, against and beneath the infrastructures underlying them, and welcomes work on the (dis-)continuous histories of technological exploration of computational infrastructures over the last 30 years or longer, and on how current artistic, theoretical and technical explorations repeat or break from them. WHO: PhD candidates and early-career researchers, with submissions welcome from researchers inside and outside the academy, including artists, designers and curators. PARTICIPANT EXPECTATIONS: participants write a paper and present it at the workshop, and on the basis of that peer review shorten the research into a newspaper-length article for distribution at the transmediale festival. The workshop is led by scholars from DARC at Aarhus University. DEADLINE: 28 August 2026, 23:59 CEST, via the application link on the call page. Queries to Pablo Velasco, pvelasco[at]cc.au.dk.
The Horizon Fellowship places technologists, scientists and researchers with emerging-technology expertise into US federal agencies, congressional offices and Washington DC think tanks to work on AI, biotechnology and emerging-technology policy. The 2027 cohort runs three tiers: junior fellows (under three years of experience, no PhD) at a USD 78,000 base for six months; fellows (three or more years of relevant experience, or a PhD) at USD 130,000 for twelve months; and a newly added senior fellow tier (twelve or more years of experience with a leadership background) at USD 190,000 or more for twelve months. All tiers receive a USD 17,000 benefits stipend, a USD 3,000 professional development stipend and up to USD 3,000 in relocation assistance, and all are potentially renewable. The programme year begins with part-time remote training from January to March 2027, followed by matching in April and May and placement from spring or summer 2027 onward. Most offers are expected by the end of November 2026. Placements run up to two years in total: fellows and senior fellows take a one-year term at a federal agency, congressional office or think tank with the option to renew for a second one-year term, potentially on a different institutional track, while junior think tank fellows take a six-month term renewable once. Horizon reports that nearly all finalists across its five cohorts to date have secured a placement at a host organization. WHAT THE FELLOWSHIP INCLUDES: dedicated training on the policy landscape and applied skills such as writing and policy analysis; a tight-knit peer community; introductions to host offices central to US emerging technology policy; and mentorship from senior leaders and the broader Horizon community. ELIGIBILITY: no prior policy experience or specific degree is required, and Horizon explicitly welcomes candidates from diverse backgrounds; fellows must reside in the Washington DC area for the placement. WORK AUTHORISATION: applicants need pre-existing work authorisation that does not depend on employer sponsorship. Non-US citizens on E-3, TN or H-1B1 visas may qualify case by case; F-1 OPT during the first year after graduation is acceptable but STEM OPT extensions are not; green card holders are eligible but may face restrictions on executive branch roles. Junior fellows have no specific degree requirement and may apply while finishing college, but are limited to the think tank track. SELECTION PROCESS (September to November 2026): application decision, then a paid work sample of up to 6 hours, a 20-minute screening interview, a 40-minute technical interview and a 25-minute final interview. Applications close 30 August 2026 at 11:59pm Anywhere on Earth.
Human Rights Watch is hiring an AI Strategy and Solutions Lead into its Operations Department to build the foundation for responsible AI adoption across the organisation. The role deliberately sits between strategy and execution: assessing organisational readiness, mapping partnerships and governance, and also prototyping and shipping working systems. THE WORK, in four strands. (1) STRATEGY AND READINESS: with the CIO and the Information Security and Technology (IST) leadership team, build an AI strategy covering data foundation requirements, infrastructure gaps and organisational readiness; assess current state and shape a roadmap balancing quick wins against long-term capacity; research what peer nonprofit, humanitarian and human rights organisations are doing and find collaboration openings. (2) FLAGSHIP SOLUTIONS: work with departments to find high-impact use cases, translate business needs into technical approaches, build and deploy AI/ML solutions with measurable value (time saved, efficiency, better outcomes), using the Microsoft AI ecosystem (Copilot Studio, Azure AI, Power Platform) among other tools, and oversee consultants and vendors to extend delivery capacity. (3) ADOPTION AND CAPACITY: balance onboarding staff onto AI tools against building custom solutions where they amplify the organisation's strengths; write training material pitched at different work areas and skill levels covering prompting and responsible use; support working groups and communities of practice; produce playbooks and guides for non-technical staff. (4) GOVERNANCE: partner with IST leadership on risk assessment frameworks, align solutions with security, infrastructure and architecture requirements, contribute to AI policy development and its operationalisation, set standards for evaluating proofs of concept and MVPs, and engage business units so outcomes are co-owned; all of it balanced against data quality, security and resource constraints and HRW's mission as a human rights organisation. WHO THEY WANT: a bachelor's degree or equivalent work experience in computer science, data science, information systems, mathematics, AI/ML implementation or a related technical field; a minimum of seven years' relevant experience including two to three years on AI and two to three years implementing AI/ML in production; experience contributing to technology strategy and assessing organisational readiness for new technology; strong Python and ML library/framework skills; hands-on predictive modelling, data science and statistical analysis; experience with low-code/no-code AI platforms such as Copilot Studio or Power Platform; a track record of scoping and delivering under resource constraints; a grasp of both generative AI and traditional ML; and the ability to explain technical concepts plainly to mixed audiences, engage stakeholders and lead technology adoption as an internal partner to business teams. PREFERRED: nonprofit, humanitarian or human rights experience; data privacy, security and ethical AI knowledge; familiarity with responsible AI frameworks and governance models; vendor and consultant management; and a background in politically sensitive or high-risk environments. APPLY: through the HRW job portal with a cover letter and CV, preferably as PDFs, by 30 August 2026. Only complete applications are reviewed and only shortlisted candidates are contacted; no calls or email enquiries. HRW states that it does not discriminate on the basis of disability, age, gender identity and expression, national origin, race and ethnicity, religious beliefs, sexual orientation or criminal record.
Touch Some Grass is an open-call poster exhibition responding to the growing presence of AI-generated imagery and its impact on creative labour, authorship and visual culture. At a time when images can be produced instantly and consumed just as quickly, the exhibition reclaims the poster as a space for ideas, intention and public conversation, inviting artists to make work that questions, challenges or reflects on this changing landscape. As the organisers put it, the question is no longer whether machines can make images, they can; the question is what kind of world we build when speed matters more than thought. NOTE: this call offers exhibition and visibility only, with no artist fee or stipend. SUBMISSION GUIDELINES: artwork size A3 (297 x 420 mm), portrait orientation, 300 DPI, CMYK recommended, print-quality PDF preferred, maximum file size 1 GB, one submission per artist. CAP: participation is limited to 60 artists and the submission form closes automatically once all slots are filled, so the effective deadline may fall well before 31 August 2026 (10 of 60 slots were taken at the time of listing). Inato Gallery is an artist-run space in Iligan City, Philippines, which began as a solo street exhibition by KaraTula and now runs concept-driven solo and group shows, open mics, talks and workshops built on openness, participation and the belief that art should exist where people already are.
The European AI & Society Fund is launching Building Narrative Power, a EUR 2 million funding and learning programme to counter Big Tech's dominance over how AI is understood, and to help people and communities shape new ways of thinking and talking about the technology they want. The Fund's strategy is to build the power of the public interest community to shape the direction of AI in Europe; narratives are central to creating the conditions for a greater say over how AI is adopted and used. THREE THEMES where research shows the public already feels strongly: (1) DATA CENTRES: as the EU aims to triple data centre numbers in the next five to seven years, pollution, energy prices and compromised grid access have mobilised communities across the political spectrum; proposals should explore narrative questions that move communities beyond resistance towards engagement in shaping data centre expansion. (2) AI AND JOBS: 57% of the public in a recent study expressed concern about how AI will change employment, more pronounced among lower-income people, and the concern is not only displacement but the quality and nature of work, autonomy and dignity; proposals should move the debate from generalised anxiety into a mobilising issue that shapes the future of work. (3) YOUNG PEOPLE: beyond the strident social media ban debate, young people encounter AI in education, friendships and future careers, and their wider views are rarely heard; proposals should build narratives that help young people shape the technologies that will influence their lives. LEARNING PROGRAMME: grantees join a community of practice, sharing experiences across issues and places, supporting each other through delivery and reflecting on which strategies and stories have most salience, with findings feeding back to the Fund and other funders. ABOUT THE FUNDER: the European AI & Society Fund is a pooled fund with 18 funding partners that mobilises a public interest community in Europe to fight for AI that serves people, society and planet; it takes no EU, government or Big Tech money. ELIGIBILITY: this call is only open to organisations in Austria, Belgium, Bulgaria, Croatia, the Republic of Cyprus, Czechia, Denmark, Estonia, Finland, France, Germany, Greece, Hungary, Iceland, Ireland, Italy, Latvia, Liechtenstein, Lithuania, Luxembourg, Malta, the Netherlands, Norway, Poland, Portugal, Romania, Slovakia, Slovenia, Spain, Sweden and Switzerland; applications from the UK, the Western Balkans and other non-European countries are not accepted. Guidance for applicants and a full application-form preview are published so you can prepare a draft before submitting. HOW TO APPLY: create an account on the Fund's application platform Optimy (link in the guidance documents) and submit the first-stage application before 31 August 2026. KEY DATES: call launched 16 July 2026; information sessions 22 July, 29 July and 24 August 2026 (11:00 to 12:00 CET); first-stage deadline 31 August 2026; unsuccessful applicants notified 16 October 2026; shortlisted applicants invited to submit a detailed proposal on 19 October 2026, due 16 November 2026; funding decisions notified 14 December 2026; grants start February 2027. Questions to info@europeanaifund.org; an FAQ covering eligibility, coalitions, non-profit status and what the grant can be spent on is published at the bottom of the programme page.
Grain is a cookbook of critical and creative AI literacy exercises for educators in film and the wider audiovisual and creative field, co-produced by AIxDESIGN and the Netherlands Film Academy and led by Nadia Piet. It publishes in March/April 2027, digital and print, under CC BY-SA 4.0. The current open call is to propose a recipe by the end of August 2026 and join as a launch contributor. WHAT THEY WANT: tried-and-tested group exercises you have actually run with people in teaching or workshops, specifically in the context of film, audiovisual work, time-based media and storytelling. Hands-on or analog, 15 minutes or a full day, beginner or advanced; the one hard requirement is that another educator can read your recipe and run it themselves. What that might look like: hands-on experiments with specific AI tools (image generators, chatbots, code assistants, video models); analog activities that reveal computational logic through embodied practice; critical readings, watchings or listenings with structured discussion; provocations, games or role-plays that get at how AI sees, thinks or makes; and exercises that help people work out how they want to relate to AI in their own practice. WHAT MAKES A GOOD RECIPE: it helps a room of people think together and trusts the room to be smart given the right container; it emerges from real practice rather than an untested idea; it is specific about what participants do, step by step, so a stranger could pick it up and adapt it to their group, time, tools or context; it has at least one designed social moment where the thinking happens between people rather than privately; it is about something larger than technique (how a machine sees an image, where your values bump up against what the tools want from you, what a film's rhythm looks like laid out in space instead of time); it takes a position without telling people what to think; and it points beyond itself with references, reading and reference projects. Published example recipes include Image Forgery: Archival Mimesis with AI (Artefacto), Machine Ways of Seeing (Nadia Piet), AI Film Now: A Close Watching Exercise (Nadia Piet) and Outside the Timeline: Spatial Grid + Visual Clustering (Artefacto), alongside a root recipe template. WHAT LAUNCH CONTRIBUTORS GET: your recipe in the first edition, print and online, credited to you; a print copy when it lands and an invite to the launch event in Amsterdam in spring 2027; support turning your exercise into a recipe, via Recipe Press (their LLM helper), the cook-along, or directly with the team; and a need-based fee of EUR 150 per recipe for freelance and independent practitioners, which you request in the proposal form. After this round contributions stay open but web-only: no fee, no dedicated support, no print. COOK-ALONG: one live 30-minute online session on Wednesday 12 August 2026 at 16:00 CEST, where you bring a rough exercise and turn it into a recipe on the spot with Recipe Press (sign up at luma.com/84w148in). HOW IT WORKS: (1) propose your recipe through the proposal form (who you are, what the exercise is, where you have run it, and whether you are requesting the fee); a short pitch is enough, and if it is already written you can just link your draft; (2) they reply on a rolling basis, with one reply covering both whether it is a match for the first edition and the fee if you requested it, so the earlier you send the earlier you hear back; (3) once it is a yes, write it up using the structure on the Contribute a Recipe page, with Recipe Press or at the cook-along. TIMELINE: proposals open until the end of August 2026, reviewed as they come in; accepted recipes due in full by mid-September 2026; Grain publishes March/April 2027. Questions to hello@aixdesign.co.
AIxDESIGN and the KB (Koninklijke Bibliotheek, the National Library of the Netherlands) are building an exhibition and public programme about AI, made from the national collection, at the KB in The Hague. Nadia Piet is Creative Lead with the AIxDESIGN community, co-presented with the KB. The premise: long before neural nets, libraries were the machines we built to organise human knowledge, and the show uses the KB's collection to look at what AI is, where it comes from, and how it connects to the older story of how we store, sort, and search what we know. Part of the show lives in the KB Atelier, with talks, workshops, and screenings alongside, intended as AI literacy you can wander into. WHAT THEY WANT: work that engages AI in some way (critical, creative, hands-on, or playful); bonus if it also speaks to archives, libraries, or how we store and sort knowledge, though you do not need to touch the KB's collection to take part. You may pitch to more than one track: (A) SOMETHING FOR THE OPENING NIGHT, a performance, talk, screening, tour, ritual, booth, game, or other experience for the November 2026 opening, fee EUR 1,500 to 2,000; (B) A SESSION IN THE EVENTS PROGRAMME, a workshop or other session once during the three-month run, around 2.5 to 4 hours, fee EUR 2,000 to 2,500; (C) AN (ART)WORK FOR THE EXHIBITION, existing work in any format (installation, sculpture, zine stand, a corner of your own) sturdy enough for a public area with daily traffic, honorarium EUR 200 to 400 including licensing and production; (D) SOMETHING ELSE ENTIRELY, pitch it anyway. WHO: artists, designers, researchers, technologists, performers, musicians, archivists, or educators working around AI, data, libraries, or cultural heritage; no finished idea or polished CV needed, just a clear, specific pitch. You can be based anywhere and they especially welcome people outside the Netherlands; the show is in-person in The Hague and the fee stays the same wherever you are coming from. HOW TO APPLY: send a short pitch through the form (which track, your idea in a paragraph or two, roughly what it would need, and a link to your work); takes about five minutes, no full application required. Deadline: end of August 2026. Questions to hello@aixdesign.co.
The ZwillGen Fellowship is a 12-month position for recent law-school graduates and early-career attorneys interested in working at the intersection of law, technology, and public policy. Fellows work alongside attorneys on matters involving privacy, cybersecurity, artificial intelligence, online platforms, surveillance, digital advertising, fintech, online safety, and technology regulation, contributing to client counseling, investigations, litigation, and emerging regulatory matters, as well as research, writing, and speaking. The program includes a reduced billable target (1,250 hours) to make room for thought leadership and engagement with the technology policy, privacy, and civil liberties community. ELIGIBILITY: must have graduated from law school before the program starts; must have taken or plan to take a bar exam near the start of the Fellowship; must be able to work out of the DC office and be eligible to apply for admission to the DC Bar. Candidates primarily interested in traditional IP practice may find it less aligned. A small number of applicants may also be considered for the Hannah Schaller Memorial Skylark Fellowship (privacy-law focus, with a reduced 1,100-hour billable target for additional community engagement). Following the Fellowship, fellows may be invited into a full-time attorney role or supported in pursuing other technology law and policy opportunities. APPLY with a resume, law-school transcript, writing sample, and cover letter by 31 August 2026.
The Digital Democracy Centre (DDC) at the University of Southern Denmark, with support from TrygFonden, is calling for applications for its 2027 fellowship cohort. The program fosters collaboration between researchers, journalists, and policymakers exploring the impact of digital technology and AI on media, politics, and democracy. Fellows collaborate with a self-selected DDC mentor to generate new knowledge, contribute to public debate, develop networks, and produce tangible outreach outputs such as policy briefs, educational packages, and talks. THREE FELLOWSHIP TYPES: (1) academic fellows researching digital technologies and democracy; (2) journalist fellows covering digital technology and democracy; (3) policy fellows contributing to the regulatory debate. Target group: researchers, journalism graduates, and policymakers in research/educational institutions, media organizations, ministries, NGOs, etc. PRIORITIES: projects with clear relevance to Denmark or Danish society (including comparative/international projects yielding insights relevant to a Danish context); in 2027, projects related to DDC's Hybrid Warfare Initiative (societal resilience to new forms of international conflict, with a focus on digital technologies and AI) are also prioritized. EXPECTATIONS: applicants must secure confirmation of mentorship from their desired DDC mentor BEFORE applying (applications without a confirmed mentor are automatically rejected; PhD students and research assistants are generally not eligible as mentors, and Sten Schaumburg-Muller is unavailable as a mentor for the 2027 call); applicants must also coordinate support from their current supervisors/managers. Fellows engage in DDC activities, seminars/workshops, and public outreach, and submit a short report within two months of finishing. APPLICATION (in English): preliminary project title, fellowship type, brief project description (max 1000 chars), detailed description and timeline (max 2400 chars, including duration, confirmed mentor, and remote/in-office work plan), description of activities and collaboration with external practice partners (max 1000 chars), expected product/impact/output (max 500 chars), a CV (max 5 pages), and an overview of the posts for which financial support is sought. Submit the application form (PDF or Word) plus CV as PDFs to ddc@sdu.dk with subject line 'DDC / TrygFonden Fellowship application'. Deadline 1 September 2026; decisions communicated no later than 1 October 2026.
Since 2020 the Prototype Fund Switzerland has strengthened the common good with open-source, public-interest technology, supporting interdisciplinary teams to develop and test early-stage projects that show how open source, data and digital technologies such as AI can create real societal value. The 2026-2027 edition focuses on Responsible and Sustainable AI: building the path toward a more responsible and sustainable AI, tech and digitalization landscape, open source and beyond. THIS YEAR'S FOCUS: projects that reduce resource use, energy consumption or emissions; strengthen accountability, transparency and governance; and contribute to resilient and trustworthy digital systems, including sustainable digitalization and digital sufficiency, public-interest AI and data infrastructures, and tools for governance, oversight or digital sovereignty. PUBLIC INTEREST TECH areas span sustainability and climate, participation and civic tech, education and literacy, health and inclusion, and open knowledge and data. WHO THEY FUND: individuals and small teams open to open source, AI and iterative experimentation, eager to combine technical capability with societal and political awareness, willing to share learnings, motivated to work on pressing societal problems through technology, with an early-stage but well-thought-out concept, and eligible to work in Switzerland. SUPPORT: up to CHF 50,000 in funding, a 4-month prototyping phase, coaching, mentoring and workshops, and access to a strong network across tech, policy and society. Teams are strongly encouraged (though not required) to publish code, data or components under open-source licenses and to document their approach. TIMELINE: applications 21 July to 7 September 2026; rolling jury selection August-September 2026; prototyping mid-October 2026 to mid-February 2027; outputs and learnings November 2026 to July 2027. Questions: info@prototypefund.ch.
Híbrides is a biennial of artistic research co-organised by the Axolot.cat collective and the DARTS (Design, Art, Technology and Society) research group at the UOC in Barcelona. The second edition, "Small Embodied Data," centres on the concept of classification -- approached through machine learning, data science, algorithmic systems, and questions of who designs classifiers, under what ethical codes, and to what end. The biennial brings classification into dialogue with its previous themes of Small Data (community-defined, intentional data collection and analysis) and Embodiment (bodily experience as a foundation of situated, sensory knowledge). Both academic and non-academic contributions are welcome. FORMATS ACCEPTED: Oral Presentations (20 minutes; artistic practice, artistic research, or academic research projects; may be video-recorded for a public archive and developed into short papers); Performances (20-30 minutes; must engage with the physical body, stage, and technology; complete works, including previously presented work; takes place in a multipurpose space with technical support). SUBMISSION REQUIREMENTS: online form with abstract (max 1,000 characters), professional biography (max 1,000 characters), and link to representative materials (3-5 images; video and technical rider required for performances). HONORARIA: oral presentations €100; performances €400; invoice required in both cases. Participants must follow the event's code of conduct. Event runs 26-28 November 2026 at UOC Poblenou campus and Convent de Sant Agustí, Barcelona. Deadline: 7 September 2026. Submit at https://hibrides.axolot.cat/opencall.html.
Hiding in Plain Site is an international public art short film festival accepting film, video art, XR, and explicitly AI-generated or AI-assisted moving image works of 2 to 20 minutes, screened in public space. Free to submit; selected filmmakers receive a modest stipend whose amount is revenue-dependent and not yet confirmed. Deadline 10 September 2026.
Horizon Europe Cluster 2 funds research on culture, creativity, and inclusive society. The 2026 call (HORIZON-CL2-2026-01-HERITAGE, launched May 12, 2026) covers 7 topics with an indicative budget of €85.5 million. Topic HERITAGE-04 is explicitly titled 'Towards a fair and transparent market for cultural and creative content in the era of generative AI,' covering: impacts of generative AI on creative industries employment; upskilling needs for creative workers; policy and regulatory frameworks; fair remuneration and transparency. Highly relevant for research into AI and cultural/creative industries, artists' rights in the AI era, and sociotechnical studies of AI and culture. ELIGIBILITY: consortium-based applications (minimum 3 organisations from 3 different EU/associated countries are typical for Horizon RIA grants; check call conditions on the Funding & Tenders portal for specific requirements). Deadline: 23 September 2026. Apply via the EU Funding & Tenders portal (ec.europa.eu/info/funding-tenders).
Call for short films (1-10 minutes) involving human/AI hybrid creation, from the We Are Human Foundation in Paris. Three cash awards: Grand Prix EUR 5,000, Best Screenplay EUR 3,000, and an Ethics Award EUR 2,000. ELIGIBILITY: international, applicants 18+; films must have been completed after 1 June 2025 and involve human/AI hybrid creation; a mandatory Ethics Notebook accompanies each submission. No submission fee. Festival screening at Forum des images, Paris, on 24 November 2026. Apply via FilmFreeway; deadline 30 September 2026.
The Wilson Center, the congressionally chartered nonpartisan policy research institution in Washington DC, awards 10 to 20 research fellowships for its 2027-2028 class. Fellows conduct research and write in their areas of interest while interacting with policymakers in Washington and with other fellows. The Center accepts policy-relevant, nonpartisan proposals that address key foreign policy challenges confronting the United States and the world, and projects must align with at least one of the Center's four scholarship pillars: strategic competition, economic statecraft, TECHNOLOGY AND INNOVATION, and regional scholarship. The technology and innovation pillar is the relevant route for AI governance, emerging technology and digital policy projects, which the Center also supports through its Science and Technology Innovation Program. Fellows work remotely for the duration of the fellowship; the Center covers travel costs for two required trips to Washington DC at the beginning and the conclusion of the term. The stipend is USD 10,000 per month for the twelve-month term, which runs 1 June 2027 to 30 May 2028. ELIGIBILITY: academic candidates need a PhD in hand and either a published book or monograph beyond the dissertation, or a substantial body of research articles (five or more) in reputable peer-reviewed journals. Non-academic applicants (practitioners, journalists, public intellectuals) need at least ten years of relevant experience and a record of substantial policy impact. There is no citizenship restriction, but English proficiency is required. Applications, including letters of recommendation, must be submitted online via Interfolio by 11:59pm ET on 30 September 2026. Questions: fellowships@wilsoncenter.org. NOTE: this is an open-topic foreign policy fellowship, not an AI-specific one; the technology angle depends on how you scope the proposal.
Waag Futurelab invites artists, theatre makers, choreographers, dancers, game designers and other creative practitioners to engage directly with seven experimental generative AI Enablers being co-created with technologists and dance, theatre and gaming professionals. The call is an invitation into the development process: selected artists get access to the technologies, datasets, interfaces and assumptions that shape these systems while they are still being built. Rather than testing whether the systems work, artists are asked to interrogate why they work the way they do, who they are designed for, what forms of creativity they enable or constrain, and what assumptions about culture, intelligence, authorship, labour, bodies, audiences and society are embedded in them. Waag is looking for individuals and collectives willing to move beyond adoption into investigation, and welcomes proposals that are speculative, critical, poetic, provocative, humorous, performative, constructive or disruptive, including work that challenges, misuses, reimagines, hacks, subverts or plays with the tools, exposes hidden biases, questions environmental costs, or imagines alternative technological futures. THE SEVEN AI ENABLERS, each requiring an on-site visit to its technology partner: the Generic Procedural Content Generation Enabler (text-to-3D assets and avatars, CERTH, Greece); the Data-Driven Cultural Content Adaptation Enabler (audience-aware creative content ideation with ranked candidate ideas, CERTH, Greece); the Cross-Domain Audience Engagement Analytics Enabler (turning heterogeneous cultural participation data into interpretable engagement signals, The Audience Agency, Warrington, UK); the Dance Style Analysis and Generation Enabler (dance video into virtual 3D choreography, dissected into body, effort, shape and space components, Centre of Excellence for Research & Innovation, Cyprus); the Immersive Design Integration Enabler for Dance & Theatre (NLP-driven XR environments with speech-to-3D scene editing, Cyprus); the Music-to-Movement Translation Enabler (audio into a time-aligned feature set spanning genre classification, emotional valence-arousal analysis, rhythm extraction and latent-space projection, feeding motion generation, spatial audio and visual choreography, Aristotle University of Thessaloniki, Greece); and the Adaptive Narrative Enabler for Games & Theatre (an evolutionary AI trained on creator-written story elements that generates infinite story variations, explicitly rejecting LLMs because of their dependence on copyrighted or public-domain corpora that encode 19th and early 20th century sexism, racism and imperialism, CAUSA Creation, Austria). ELIGIBILITY: artists, designers, creative technologists, theatre makers, dancers, choreographers and game designers, as individuals or collectives; applicants must hold a European VAT number. An info session with live demonstrations of the AI Enablers runs 31 July 2026, 10:00 to 11:30 CEST (meeting ID 995 7016 2341). Applications go through Homerun using the official Application Template, uploaded as a PDF; read the Guide for Applicants, which contains the full Enabler descriptions, before applying. Questions: zoenie at waag dot org. The AI Enablers and this call are part of the Horizon Europe project HAMLET, which aims to make generative AI accessible to cultural organisations of all sizes. Closes 30 September 2026, 23:59 CEST. Free to apply.
The AAAS Science & Technology Policy Fellowships place scientists and engineers inside US federal agencies, congressional offices and the judicial branch for a year of hands-on policymaking, contributing technical knowledge and analytical skills to the federal policy process. AI, data and computing policy placements are among the largest tracks, and AAAS runs a dedicated AI legislative branch cohort that places AI experts in congressional offices at GS-12 Step 10 on the DC locality scale. ELIGIBILITY: US citizens holding a PhD or equivalent doctoral degree in any science or engineering field, or a master's degree in engineering plus three years of post-degree professional experience; some tracks accept a bachelor's plus five years. Relevant expertise includes AI, machine learning, computer science, robotics and data science. Fellows must reside in DC, Maryland or northern Virginia during the fellowship year. TIMELINE: the application portal opens 1 August 2026 and closes 1 November 2026 at 11:59pm ET, with eligibility notification in late November, interview notification January to February 2027, semi-finalist interviews in March, placement offers by early May, and the fellowship year running 1 September 2027 to 31 August 2028. Apply through fellowshipapp.aaas.org.
The Omidyar Network Tech Journalism Fund offers rolling grants of $5,000-$25,000 for technology journalism projects. Relevant for journalists and media organizations investigating AI, tech policy, and AI's impacts on culture, society, and creative industries. Applications are accepted on a rolling basis through December 31, 2026 — one of the few currently open opportunities with a confirmed near-term deadline. Apply via the application form on omidyar.com.
The New Media Film Festival in Los Angeles champions stories told with new technology and was among the first festivals to screen AI-generated content; it has a dedicated AR / MR / AI / TECH category. Open worldwide in all formats. Deadlines: 6 December 2026 (USD 80), 6 February 2027 (USD 85), final 6 April 2027 (USD 95).
AIxDESIGN gets more requests for speakers and facilitators than it can handle in-house, so rather than turning them down it passes them through to a curated pool of critical, creative and thoughtful voices on AI. WHAT IT IS: a Speakers Collective whose members are connected with opportunities as they come in, including talks, workshops, panels and custom sessions for organisations, festivals, institutions and events. AIxDESIGN makes considered matches, briefs speakers on the gig and supports where needed; in return speakers credit AIxDESIGN publicly and contribute a small coordination fee or donation back to the ecosystem. Opportunities are posted in a private Speakers Collective Slack channel, where you raise your hand for what interests you and skip what does not. There is no exclusivity requirement. WHO THEY ARE LOOKING FOR: thinkers, makers, educators, researchers, artists and practitioners with honest, grounded perspectives on AI, who care what it means for society, culture and creative practice rather than only what it can do. They are particularly drawn to people working at the edges, questioning dominant AI narratives and exploring what critical, slow, speculative or decolonial approaches look like in practice, or making work that opens new imaginaries rather than reinforcing existing ones; practices at the intersection of AI and art, design, sound, moving image, writing, education, policy or community are a good fit. You need something real to say and some experience saying it (a past talk, workshop, panel or teaching context that can be pointed to), and alignment with AIxDESIGN's values. BONUS POINTS: if you have worked with AIxDESIGN before and are comfortable speaking about the shared project, or if you are based in or regularly in the Netherlands. PROBABLY NOT FOR YOU: if you have no prior speaking experience, or if your perspective clashes with what they stand for. HOW TO APPLY: a roughly five-minute form (Tally). Applications are accepted on a rolling basis and reviewed about every two months; if it is a fit, they invite you into the private channel where opportunities are shared.
Robot foundation models (pi-zero, OpenVLA, GR00T and others) are approaching widespread deployment on the back of more than USD 18 billion in funding, and almost nobody is working on their safety. The Physical AI Safety Institute, a new nonprofit founded by Bear Haon and Kaylene Stocking and seeded by a grant from BlueDot Impact, describes itself as a capacity-building organisation growing the community that develops techniques to interpret, align and control robot foundation models. Its research fellowship runs in 12-week cohorts: each fellow works within a stream led by a mentor on a concrete problem in physical AI safety, with mentors drawn from researchers with track records in robot foundation models, mechanistic interpretability, safe control or adversarial robustness for embodied systems. Each cohort ends with a research symposium where fellows present their work. STREAMS: mechanistic interpretability (sparse autoencoders and activation patching for VLA and WAM architectures); adversarial robustness (jailbreaking and prompt injection in the embodied action space); runtime monitoring (detecting when a policy operates outside its competence envelope); sim-to-real safety transfer (do safety behaviours survive deployment on physical hardware); and robot security (cyberattack vectors, data poisoning and supply chain for deployed robot models). WHO SHOULD APPLY: researchers with strong technical foundations in robotics or machine learning and a demonstrated interest in safety; prior safety experience is not required. PROCESS: this stage is an expression of interest, not a full application. The form takes about two minutes and asks you to pick your stream and preferred cohort dates; concrete details including stipend figures and application deadlines are shared through that channel, and neither is published on the site as of listing. PAISI also publishes seven free curriculum modules running from classical robot safety through VLA interpretability, adversarial robustness and runtime monitoring to deployment evaluation, plus a white paper, The Case for Physical AI Safety, at paisi.ai/the-case-for-physical-ai-safety. Contact: bear@paisi.ai.
Playlab is a nonprofit building public infrastructure for educational AI, equipping communities to create AI tools that reflect their own needs and values; more than 60,000 educators have published apps on the platform. Its premise is that AI is a new design material that should be shaped by many, and that educators should be designers of AI systems rather than end users of someone else's. THE ROLE: visual communication designers own the brand and the surfaces where people encounter and build with it, shaping how educators learn about AI and gain the confidence to tinker with it, and making sophisticated ideas legible without flattening them. Work spans brand expression, storytelling, data and information design, motion, interface and print, taken from concept through produced work and shipped as reusable systems. The job includes getting close to the people using Playlab and the contexts they work in, and helping collect and share their stories. WHAT THEY LOOK FOR: craft that is production-grade with a clear point of view and good judgement about when to ship fast versus slow down; narrative sense, meaning well-reasoned positions on how Playlab tells its story and the ability to turn complex systems into intuitive visuals that stay true to their depth; initiative on fuzzy problems, from defining the space to shipped work, thinking ahead to dependencies and second-order questions; technical fluency across print, web and motion, building with and on AI as a creative material and curating tooling and workflows; and taste, a distinct visual voice with the restraint to leave a rough edge when it serves the work. FIT: at least 5 years' experience with a strong graphic design foundation, a home base in visual communication (brand, editorial, campaign, web) and a track record in at least one other design discipline; able to lead design on owned projects while collaborating across the team; broad knowledge of creative tools and workflows including AI-leveraged ones; genuine interest in learning environments and the people in them; and strength as a producer comfortable owning the last mile to production. BONUS: education or edtech experience, designing creative tech, and a love of creative side quests. The posting explicitly encourages applications from people who do not meet every listed qualification. Culturally the organisation describes itself as sitting between a startup and an R&D lab, with nonprofit governance intended to keep incentives tied to teaching and learning rather than growth.
Playlab is a nonprofit building public infrastructure for educational AI, with more than 60,000 educators publishing apps on its platform. Designers there are responsible first for researching, prototyping and implementing interaction patterns that let educators be makers of technology rather than recipients of someone else's system, then for shaping how educators learn about AI and gain confidence to tinker with it, and finally for the storytelling that carries the work into the world. The two standing questions the role works on: how to expand and deepen what educators can create with AI, and how to lower the floor for who can create and how easily AI systems can be meaningfully deployed in real educational contexts. THE ROLE: as Senior Product Designer you own the design of Playlab's surfaces from prototype through shipped product. WHAT THEY LOOK FOR: craft that is production-grade with a clear point of view, and instincts for when to ship fast versus slow down; product sense grounded in staying close to the problems Playlab is solving, with the ability to hold unresolved questions and keep moving; initiative on fuzzy goals, defining the design space, gathering context and driving through prototyping to a shipped solution while bringing leadership along; technical fluency, building with and on AI, judging what is worth building, and curating custom tooling; and taste across visual, interaction and communication design. FIT: at least 5 years as a strong design generalist with a home base in product design and a track record in another design discipline; able to lead design on owned projects while collaborating across the team; broad knowledge of creative and AI-leveraged tools; interest in learning environments and the people who create them; and comfort getting technical with engineers, possibly pushing code. BONUS: education or edtech experience, designing creative tech, and a love of creative side quests. The posting explicitly encourages applications from people who do not meet every listed qualification. The organisation describes itself as sitting between a startup and an R&D lab, with nonprofit governance keeping incentives aligned to teaching and learning rather than growth.
A research role on Microsoft's AI Red Team, an interdisciplinary group of researchers, red teamers and engineers working to understand and reduce security risks in Microsoft's AI systems, with impact across the Phi series, Bing Copilot, Security Copilot, GitHub Copilot, Office Copilot, Windows Copilot and Azure OpenAI. The brief is agent security: how autonomous, tool-using and multi-agent systems fail, and what principled methods can evaluate and improve them. RESPONSIBILITIES: formulate and investigate research questions at the intersection of cybersecurity and agentic AI behaviour, covering security risks, failure modes and adversarial behaviour; design rigorous experiments, evaluations and benchmarks for autonomous, tool-using and multi-agent systems; develop research insights, methodologies and mitigations; and work with product teams to turn findings into practical security improvements. REQUIRED: a Bachelor's in statistics, econometrics, computer science, electrical or computer engineering or a related field plus 2+ years of related research experience, OR a Master's plus 1+ year, OR a Doctorate, OR equivalent experience. Candidates must be able to pass the Microsoft Cloud background check on hire and every two years after. PREFERRED: demonstrated research output in AI security, adversarial machine learning, agentic systems or LLM evaluation (publications, preprints, technical reports, talks, open-source tools or research artifacts); experience designing experiments, evaluations, benchmarks or methodologies for security-relevant behaviour in AI systems; comfort in an interdisciplinary team; and the ability to connect findings to practical security guidance for shipped products. Posted 29 July 2026 (job number 200045204); open for a minimum of 5 days and then until filled, so apply early. Background on the team's approach: https://www.microsoft.com/en-us/security/blog/2023/08/07/microsoft-ai-red-team-building-future-of-safer-ai/
The Responsible AI team inside Microsoft's CoreAI Platform and Tools organisation is hiring at two levels (Applied Scientist II and Senior Applied Scientist) to build the RAI service components used across Microsoft's AI products by hundreds of millions of people. RESPONSIBILITIES: supervised fine-tuning of LLMs including reinforcement learning from human feedback, with thorough evaluation; working with the tech lead on the design and implementation of agent adversarial evaluations and safety mitigations so that agent creators can identify risks and apply safeguards; and collaborating with the RAI engineering team to ship those mitigations at scale, refine them against operational performance and resolve end-to-end issues. REQUIRED: a Bachelor's in computer science, computational linguistics or a related field plus 2+ years of related experience, OR a Master's plus 1+ year, OR a Doctorate, OR equivalent experience. Candidates must pass the Microsoft Cloud background check on hire and every two years after. PREFERRED: a Master's plus 6+ years or a Doctorate plus 3+ years; 2 to 3+ years owning feature design, model development, evaluation and deployment; strong machine learning background especially NLP and deep learning; language model training and evaluation experience; Responsible AI research experience; Python and C# with PyTorch and Triton; large-scale data processing; unit testing, TDD and DevOps practice; platform and API design, debugging and data analysis; full-stack (client to service) development is a plus. Posted 10 July 2026 (job number 200017828); open for a minimum of 5 days and then until filled.
AIXI Labs is a small London-based non-profit doing AI safety research grounded in algorithmic information theory. Its stated position is that most AI research today is either mathematically clean but narrow, or practical but too opaque for rigorous safety analysis; the lab targets a third category, methods general enough to describe powerful agents and provable enough to support real safety claims, using AIXI (the theoretical model of unbounded artificial superintelligence) and then porting the strongest of those ideas to modern LLM-based agents. The fellowship is a quarterly, rolling programme for researchers who want to work on this agenda under close mentorship from the core team: Cole Wyeth (Founder and Executive Director), Marcus Hutter (Research Director) and Aram Ebtekar (Founding Research Scientist). Fellows pursue a focused research project over roughly a quarter and receive a monthly stipend plus a compute budget for machine-learning work. The fellowship is fully remote, so fellows may work from anywhere in the world or from the London base. RESPONSIBILITIES: pursue a self-directed research project aligned with the lab's algorithmic-information-theoretic approach to AI safety, spanning anything from foundational theory to empirical tests with LLM-based agents; shape the project direction around your own strengths and interests, as the lab supports a range of work within this framework; and contribute papers, blog posts or other written output communicating results to the broader AI safety and ML research communities. MINIMUM QUALIFICATIONS: a research background at the level of a current PhD student (for primarily empirical projects, a master's-level command of the relevant mathematics paired with substantial industry or applied research experience can also qualify); solid grounding in probability, statistics, information theory and reinforcement learning, sufficient to engage seriously with AIXI-style theoretical models (prior expertise in algorithmic information theory or AIXI is NOT required); strong verbal and written communication skills; genuine motivation to reduce existential risk from advanced AI and enthusiasm for the lab's kind of work. PREFERRED QUALIFICATIONS: publications at top ML, AI or theory venues such as NeurIPS, ICML, ICLR, ALT or COLT, though the lab values the strength of your best work over volume; background in algorithmic information theory, computability theory, learning theory or related theoretical computer science; experience fine-tuning, evaluating or red-teaming LLM-based and/or RL agents; prior engagement with the UAI/AIXI community or related fields, including members of the lab's online reading groups. APPLICATIONS are reviewed on a rolling basis and previous applicants remain eligible for later rounds.
The Institute for Advanced Study is seeking a 2026-27 Civic Science Fellow to advance understanding of the social, democratic and scientific implications of artificial intelligence. The full-time, 18-month position is based in the Science, Technology, and Social Values Lab led by Harold F. Linder Professor of Social Science Alondra Nelson. Working closely with Professor Nelson and Lab members, the Fellow contributes to research and policy analysis and to scholarly and public-facing publications, and supports multidisciplinary, multisector initiatives at the intersection of AI, science and governance. DUTIES (75% Lab work): conduct original research on the Lab's AI-and-society agenda; synthesize scholarship on the social implications of AI, sociotechnical approaches to AI governance and the epistemological implications of AI in research; contribute policy research on governance mechanisms that strengthen democratic institutions; help organize convenings and working groups; collaborate on manuscripts, reports and grant proposals; support editorial development of publications; and develop public-facing essays, policy briefs and digital materials. The remaining 25% supports the Fellow's own independent research, network-building and full participation in the intellectual life of IAS. MINIMUM QUALIFICATIONS: a PhD in the social sciences, sciences or humanities; a strong record of scholarly research, writing and publication on the social implications of science and technology or related fields; and exceptional writing and editing skills for both academic and non-academic audiences. Preferred: specialization in science and technology studies, science and technology policy, organizational studies or qualitative methods, and demonstrated engagement with AI governance and technology policy debates. Apply with a cover letter and resume via the IAS/Paylocity posting.
Manifund's regranting program funds AI safety and global-catastrophic-risk work through a network of expert regrantors, each given an independent budget to recommend grants to people and projects they find promising. Applicants submit a project proposal on the Manifund platform; regrantors can then offer funding. GRANTS: typically USD 5,000 to 50,000, though amounts vary. ELIGIBILITY: individuals (no legal entity or institutional affiliation required) and organizations working on AI safety or reducing catastrophic risk from AI. Applications are accepted on a rolling basis; per-regrantor budget availability varies over time. Learn more at the regranting page and apply on the Manifund platform (manifund.org).
The Sony Innovation Fund (now under Sony Ventures Corporation) is Sony's corporate venture capital arm, investing in early-stage (Seed, Series A and Series B) transformational-technology startups across business, entertainment and society, including entertainment, gaming, AI, robotics, fintech, climate and blockchain. Music, entertainment and AI startups fit within its thesis, though it is not a music-only fund. Typical checks are roughly USD 100,000 to USD 5 million, alongside mentorship and access to Sony's network and R&D. It is an equity investment, not a grant. Global, with offices in the US, London/EU, Tokyo, Bangalore and Israel. Startups engage on a rolling basis via the contact and inquiry form. No application fee.
Humans in Control empowers the public to shape AI policy through grassroots organizing and bipartisan advocacy for commonsense safeguards that protect children and families. The Distributed Organizing Manager builds and supports distributed/volunteer organizing to advance the organization's advocacy goals. This is a full-time, fully remote role open to candidates anywhere in the United States (US work authorization required). Compensation is USD 65,000-80,000 plus benefits. Applications are reviewed on a rolling basis (no fixed deadline). See the listing for full responsibilities and to apply.
The CHAI Research Fellowship trains highly qualified postdoctoral researchers to advance beneficial AI, working with CHAI faculty Stuart Russell, Pieter Abbeel and Anca Dragan and collaborating with affiliate faculty at Berkeley AI Research and beyond. Fellows have broad freedom to pursue novel research toward provably beneficial AI systems in areas such as reasoning, decision making, learning, multi-agent systems and philosophical foundations, using probability theory, game theory and control theory. Current topics of interest include theoretical foundations of intelligent agent architecture, expressive formal languages for probability models (including probabilistic programming), decision making over long time scales, cumulative lifelong learning, human-in-the-loop planning and reinforcement learning, human-robot interaction, the structure of human preferences, acting on behalf of multiple humans, robust cooperation in heterogeneous multi-agent systems, and mechanism design for human-machine systems. Fellows may also lead collaborations, advise junior researchers, and teach. QUALIFICATIONS: a PhD (or about to be obtained) in a relevant technical discipline (computer science, statistics, mathematics, or theoretical economics) and a record of high-quality published research; prior work on the AI control problem is not required. International applicants are eligible (visa sponsorship provided). TO APPLY: submit via the CHAI application form a resume, a one-page statement of interest describing the research you would like to undertake, and the names and emails of two academic referees. Start date is flexible and applications are reviewed on a rolling basis (no fixed deadline). General enquiries: chai-info@berkeley.edu.
Foresight Institute grants supporting work at the intersection of AI, science and safety, including AI-for-science tools and AI-safety research. ELIGIBILITY: individuals, teams and organisations (for- and non-profit) worldwide; applicants active in the San Francisco or Berlin hubs are prioritised. Applications are reviewed on a rolling basis with a cutoff on the last day of each month. Apply via the Foresight Institute grants page.
Effective Altruism Funds' Long-Term Future Fund makes grants to individuals and small groups working to reduce risks from advanced AI and other threats to humanity's long-term future, including technical AI safety and alignment research, field-building and communication. ELIGIBILITY: individuals and small groups; not geographically restricted. Applications accepted on a rolling basis via the EA Funds application form.
Research residency at Perplexity for exceptional researchers, engineers and analysts from any quantitative or analytical discipline (physics, cognitive science, quantitative finance, theoretical mathematics, etc.) to apply their expertise to AI research challenges including agentic systems and human-AI interaction. ELIGIBILITY: no PhD or deep-learning-framework experience required; instead, strong mathematical foundations, demonstrated technical depth in a rigorous quantitative field, curiosity-driven research practice and clear communication skills. FORMAT: in-person residency paired with senior researcher mentors, with access to compute, research tools and proprietary datasets; residents are encouraged to publish at top-tier conferences and contribute to open source. Rolling start within ~8 weeks of acceptance. Applications via the Perplexity portal: resume/CV or portfolio plus a cover letter describing your interest in AI research at Perplexity; an optional research statement of project ideas is welcomed.
Rolling grant programme funding individuals at any career stage who want to pursue careers that could help reduce global catastrophic risks or otherwise improve the long-term future. Especially interested in candidates working on risks from future advances in AI and global catastrophic biological risks. Funds graduate study (master's/PhD/MPP/law school), unpaid internships, postdocs, professional certifications, online courses, independent study/upskilling, career-transition and exploration periods, and academic sabbaticals. Concrete examples Coefficient Giving lists: a senior ML engineer doing six months of independent study to investigate AI risk mitigation careers; a physics PhD doing self-guided ML interpretability work to transition into technical AI safety; a management consultant exploring how to apply their skill set to GCR; a tenured ML/CS professor taking a one-year sabbatical to contribute to AI safety or governance. Open globally; no institutional affiliation required. Looks for candidates whose funding would 'make a difference' (otherwise unable to find sufficient funding, or existing funding has restrictions). Encourages applications from women and people of color. Now subsumes the former Biosecurity Scholarship. Applications are open until further notice and assessed on a rolling basis. Free to apply.
Supports journalists producing long-form investigative reports and tech-policy researchers conducting research, analysis or evaluation that informs policy related to the Kapor Foundation's three priority areas, with an emphasis on responsible AI and tech ethics. Priority areas: CS/AI Education, Innovation, Governance. Aims to dismantle systemic inequities in the tech sector by funding researchers and investigative journalists exploring barriers and driving actionable solutions. Applications accepted on a rolling basis; first awards to be announced by 30 June 2026.
Lightweight rolling grant for individual journalists worldwide (writers, photographers, radio, film; freelance or staff) examining how AI systems are designed, sold and deployed in communities. Faster turnaround than the Pulitzer AI Accountability Fellowship: decisions in 1 to 2 weeks.
PAID 12-week online mentorship program from Belgrade Art Studio: each artist works one-on-one with a dedicated artist-mentor through weekly 1-hour Zoom/Google Meet sessions, aimed at artists at a turning point in their practice (stuck, seeking new directions, experimenting with unfamiliar media including AI). Hybrid seminar/workshop format supporting both physical and online presentation, plus curatorial thinking, writing, interviews, and documentation. Closes with an online viewing and moderated discussion of the works produced. ELIGIBILITY: open worldwide to artists and creatives across visual arts, music, theatre, dance, film, writing, design, photography, digital/interactive art, games, fashion, architecture, and curation. English-language program (ability to collaborate in English preferred but accommodations possible). EXPECTATIONS: artists must treat the program as they would an in-person residency (dedicated time for research/creation/reflection), submit in-progress material for social-media exposure, participate in an artist interview, and refrain from hate speech or discriminatory content. APPLY by emailing CV, summary, and portfolio (links accepted) to artlab@belgradeartstudio.com. Rolling intake (no deadline). Selected artists are announced and contacted by email. COST: EUR 675 to participate — this is a paid offering, not a funded residency.
The microagi Research Fellowship (Cohort 01) provides researchers with frontier-lab resources and the freedom to work on hard problems in robotics, embodied AI, and the systems around them. Fellows work in residence for six months at microagi's offices in Munich or Zurich. AMOUNT: up to $2,000,000 in compute; $10,000/month stipend. RESEARCH AREAS: (1) Long Horizon Dexterous Manipulation -- end-to-end models for complex multi-stage tasks with dexterous hands, natural language instruction following, memory across long task sequences, fine-grained physical control in unstructured environments; (2) Cross Embodiment Learning; (3) On Device Physical Intelligence; (4) Safe and Reliable Deployment. microagi's model is to deploy embodied intelligence into physical-world operations and turn what it learns there into compounding research. REFERRAL: $10,000 bonus for referring a candidate who is accepted into the cohort. No application deadline publicly stated as of June 2026 -- one cohort only, apply as soon as possible via https://www.microagi.ai/fellowship.
The Good Life Residents Program is a six-month Roosevelt Institute initiative that develops new leaders and elevates bold ideas across six economic policy priorities under its Good Life Agenda. Each topic area hosts a cohort of 3 to 4 residents plus 1 to 2 senior Topic Advisors (academics, policy leaders, organizers or government practitioners) who provide strategic and intellectual mentorship and moderate the closing event. Residents write a roughly 1,500-word original policy essay advancing a people-centered policy vision against a topic-specific prompt, take part in structured dialogue with peers, Topic Advisors and Roosevelt Think Tank directors, help prepare and appear at a culminating public event, contribute to press opportunities such as op-eds and reporter calls where applicable, and act as ambassadors for the Good Life Agenda in their professional and public networks. The current round, the AI and Finance Residents Program, is run in partnership with Data & Society and asks for bold ideas on how to shape, redirect, interrupt or make more visible and contestable the flows of capital into generative AI. Framing: breakneck investment in generative AI technology and infrastructure is distorting the stock market, profit measures, bond issuance, the pace and composition of growth, job creation and trade policy, with seven tech companies accounting for more than a third of S&P 500 market capitalisation and complex, sometimes circular financial arrangements obscuring risk. Suggested lines of inquiry include what government interventions could improve public oversight, risk allocation and direction-setting in private markets; how other investments could be coordinated and scaled if AI spending is crowding them out; what role public spending or new institutions could play in catalysing technology R&D designed to support workers, sustainability and supply chains; what democratic input mechanisms would keep state influence over capital flows less vulnerable to corruption; and how federal AI policy should address public hostility to data centre investments alongside national security and competitiveness. Compensation is USD 3,000 plus travel and accommodation for the culminating event. FLAG: Roosevelt does not publish an application deadline for the AI and Finance round on the programme page, and earlier rounds (Social Security financing, Federal Reserve powers, housing) each ran their own short application window. Check the Formstack application form linked from the programme page for the current closing date before applying.
Luminos.AI, an embedded AI-governance and legal-risk infrastructure company (built by former Fortune 500 AI governance leads and company-builders), runs a part-time fellowship for a technically inclined law student or recent graduate who wants to work at the frontier of AI agent oversight and risk evaluation. Fellows work directly with senior lawyers and engineers on the practical problem of how to test, evaluate and govern generative and agentic AI systems that change every quarter. WHAT YOU DO: research and help design rigorous testing and evaluation approaches for generative and agentic AI, translating fast-moving legal and regulatory requirements into concrete, testable standards; track the global AI regulatory landscape in real time and turn it into practical guardrails; draft frameworks, workflows and risk assessments that hold up under real deployment conditions; and partner closely with the technical team, where comfort around code, models or data is an asset. Fellows with technical skills who want to spend part of their time working directly with engineering are encouraged to do so. WHO YOU ARE: currently pursuing a JD (or a recent graduate) from a top law program; an excellent researcher and writer comfortable applying existing law to genuinely new situations; genuinely enthusiastic about AI and willing to learn the technical details. Bonus: coding ability, ML/GenAI exposure, or a background in privacy, security or data governance law. TERMS: USD 35 per hour, remote, part-time, flexible 3 to 6 month term, with direct mentorship from senior legal and technical leadership. TO APPLY: send a cover letter, resume, transcript and writing sample to careers@luminos.ai with the subject line 'Application for Luminos.AI Fall 2026 Fellowship'. Rolling review with remote interviews.