Next generation artificial intelligence (AI): explainable AI

Next generation artificial intelligence (AI): explainable AI is sponsored by Engineering and Physical Sciences Research Council (EPSRC), UK Research and Innovation. This opportunity supports mission-aligned projects and measurable outcomes.

Next generation artificial intelligence (AI): explainable AI – UKRI Funding opportunity: Next generation artificial intelligence (AI): explainable AI UK Research and Innovation, Engineering and Physical Sciences Research Council (EPSRC) 11 August 2026 9:00am UK time 20 October 2026 4:00pm UK time Apply for funding for projects focused on speculative and high-risk fundamental research with the potential to deliver high reward and a step change in the explainability of future artificial intelligence (AI) systems.

Applications are particularly welcomed from early and mid-career UK researchers. You must be based at a UK research organisation eligible for UK Research and Innovation (UKRI) funding. The full economic cost (FEC) of your project can be up to £602,500.

UKRI will fund 80% of the FEC (£482,000). Projects must start on 1 February 2027 and last for up to 24 months. This opportunity is open to organisations with standard eligibility.

Check if your organisation is eligible . Funding for this opportunity is subject to full business case approval. To apply you must meet the standard eligibility requirements for UKRI .

Applications are particularly welcome from early and mid-career researchers. The inclusion of researcher co-leads is encouraged where they have made a substantial intellectual contribution to the development of the application and will be engaged with the ensuing research.

Given the short-term nature of the grant and fixed start date, the project lead (PL) may choose to contribute a greater proportion of their time to the project than is standard. This should be justified and proportional to the activities set out in the proposal. In line with our equality, diversity and inclusion principles, joint leadership of projects will be allowed.

You must agree an administrative lead from an eligible UK research organisation to be the project lead on the application form, and their institution will be responsible for submitting the grant application. You may be involved in no more than two applications submitted to this funding opportunity. Only one of these can be as project lead.

International researchers International researchers can only apply as ‘project co-lead (international)’ as part of an applications making use of the UKRI-RCN Money Follows Cooperation agreement or the UKRI-IIASA agreement. You should include all other international collaborators (or UK partners not based at approved organisations) as project partners.

We will not accept uninvited resubmissions of projects that have been submitted to UKRI or any other funder. Find out more about EPSRC’s resubmissions policy . Equality, diversity and inclusion We are committed to achieving equality of opportunity for all funding applicants.

We encourage applications from a diverse range of researchers and organisations. We support people to work in a way that suits their personal circumstances. This includes: support for people with caring responsibilities alternative working patterns UKRI can offer disability and accessibility support for UKRI applicants and grant holders during the application and assessment process.

Demand management is not currently being applied to this funding opportunity. However, should the level of interest exceed what can be managed within the assessment process, UKRI may introduce additional limits on the number of applications that can be submitted. Further details will be clearly communicated where this is the case.

UKRI encourages organisations to support applicants in preparing well-planned, high-quality applications that are competitive for funding relative to the funding opportunity. Artificial intelligence (AI) is transforming research, industry and public services, but progress is not universal across sectors.

Increasing AI explainability could unlock a further wave of innovation with the potential to advance safety, reliability and responsible use. This could underpin the next generation of AI technologies and create new market opportunities.

We are investing up to £10 million over two years into speculative and high-risk fundamental research with the potential to deliver high reward by enabling a step change in the explainability of future AI systems. support potentially transformative, highly novel, fundamental research with the potential to deliver a step change in the explainability of future AI systems.

This would lay the foundations for future societal, economic and technological benefits, including new market opportunities for the UK strengthen the UK’s capability to design, develop and deploy explainable AI technologies broaden access to strategic AI funding, particularly for new entrants and early and mid-career researchers in the UK enable small-scale, rapid projects that generate foundational insights which can be built on in the future deliver a faster process compared to current standard funding routes, while maintaining robust decision making that is proportionate to the level of risk This opportunity is intended to stimulate creativity within the research community and give researchers time and space to test out speculative ideas.

Applicants are advised to strongly consider whether their research idea is truly speculative, high-risk and could potentially offer high reward before applying.

aim to disrupt existing areas of explainable AI research or create new ones, delivering research with the potential to lead to radically new and disruptive technologies and a step change in explainability of future AI systems adopt a risk tolerant approach utilising ways of working that encourage and enable transformative outcomes rather than incremental advances support an ambitious and capable project team with the appropriate breadth and depth of expertise required to achieve the research objectives Projects are not expected to: produce tangible outcomes, such as new AI models or spinouts, within two years succeed in all of their research goals, especially the most risky and ambitious avoid failure; approaches that do not work as intended can still generate important learning and valuable knowledge for the wider research community We encourage you to be as adventurous as possible in your thinking.

We will support creative and potentially unconventional ways of working. We particularly encourage applications from a broad and diverse range of researchers, including early- and mid-career researchers, to help capture a wider diversity of research excellence, creativity and approaches from across the community to enable more transformative outcomes. Interdisciplinary approaches are welcome.

However, we aim to support research that is working towards a demonstrable impact on AI technology and build the UK’s capacity to design and deploy explainable AI technologies. We will not fund applications that consist of majority non-technical research such as purely social science, legal, policy or ethical studies.

We will not support incremental work lacking ambition to push the boundaries of AI explainability, nor the continuation of existing projects. We expect applicants to explain why their proposed research cannot be supported via other funding routes (for example, applicant-led responsive mode or utilising the flexibility afforded by critical mass funding).

We are taking a broad, problem-focused approach to remit in this funding opportunity and intend to support applications that address the aims described above. We welcome proposals from a diverse range of research backgrounds, including mathematics, computer science, AI, and other relevant disciplines, and encourage applicants to explore how their expertise can contribute to advancing AI explainability.

Explainable AI research typically seeks to address questions including, but not limited to: how AI systems can be explained and how those explanations are understood by different audiences why AI systems produce particular outputs, behaviours, or decisions Examples of active areas of explainable AI research include, but are not limited to: output reasoning: understanding why an AI model produces a particular output through post-hoc explanations, such as text-based explanations, counterfactuals, reasoning about possibilities and probabilities, and explanations tailored to users with different levels of expertise mechanistic interpretability: understanding how the internal structure and mechanisms of AI systems give rise to their outputs and behaviours process examination: understanding the steps an AI system takes to reach a conclusion, including chain-of-thought-style reasoning, tool use, and interactions between different models or components imposed explainability: designing or constraining AI systems to make their operation more interpretable, for example through physics-informed or otherwise inherently explainable approaches uncertainty quantification: assessing and communicating the reliability, confidence, and limitations of AI outputs There is no expectation that applications will fall within any of the categories listed above.

These examples are illustrative rather than prescriptive. We particularly welcome novel and potentially paradigm-shifting approaches that challenge existing assumptions or extend beyond established areas of explainable AI research. For more information on the background of this funding opportunity, go to the ‘Additional information’ section.

We expect teams to reflect the required skills and expertise to achieve the proposed plan. As outlined under ‘Who can apply’, due to the short-term nature of the grant and fixed start date, the project lead (PL) may choose to contribute a greater proportion of their time to the project than is standard. This should be justified and proportional to the activities set out in the proposal.

Successful applicants are expected to actively engage with two annual workshops that will build a cohort of explainable AI experts, facilitate further research development, and provide information on potential future funding avenues. Project teams are expected to engage with planned monitoring and evaluation.

Full details will be confirmed with successful applicants but is expected to include a survey three months after the grants start, and reports at the end of years one and two. Joint leadership of projects will be allowed, but you must agree an administrative PL responsible for submitting the grant application.

Any proposed job-sharing arrangements should be detailed and justified as part of the ‘Applicant and team capability to deliver’ section of the application and will form part of the assessment criteria. You may be involved in no more than two applications submitted to this funding opportunity. Only one of these can be as project lead.

Teams must indicate which members will participate in the distributed peer review process, this should include the project lead and a co-lead if appropriate. As such, you can only be involved in two projects, which will help to reduce reviewer load, minimise conflicts and ensure truly speculative ideas are submitted. The maximum duration of this award is 24 months.

Projects must start on 1 February 2027. The FEC of your project can be up to £602,500. UKRI will fund 80% of the FEC.

Due to the nature of this funding, no-cost extensions arising due to non-people related reasons will not be allowed. This is a maximum amount not a target. UKRI welcomes smaller proposals as well as larger ones.

All costs must be fully justified and proportionate to the ambition of the project.

contribution to the salary of the project lead and co-leads research, technical and project management staff research consumables including compute costs travel and subsistence, including that related to cohort management activities such as the annual workshop and networking data preservation, data sharing and dissemination estates and indirect costs equipment with a value over £25,000 per item (including VAT).

Smaller items of equipment (individually under £25,000) should be in the ‘Directly incurred – other costs’ heading EPSRC’s approach to equipment funding is applicable for this funding opportunity. EPSRC standard guidance on intellectual property applies to projects supported by this funding opportunity.

Successful projects from this opportunity which show promise can seek further funding afterwards through the current routes, for example, standard mode proposals or fellowships. Supporting skills and talent We encourage you to follow the principles of the Concordat to Support the Career Development of Researchers and the Technician Commitment .

Trusted Research and Innovation (TR&I) UKRI is committed in ensuring that effective international collaboration in research and innovation takes place with integrity and within strong ethical frameworks.

Trusted Research and Innovation (TR&I) is a UKRI work programme designed to help protect all those working in our thriving and collaborative international sector by enabling partnerships to be as open as possible, and as secure as necessary. Our TR&I Principles set out UKRI’s expectations of organisations funded by UKRI in relation to due diligence for international collaboration.

As such, applicants for UKRI funding may be asked to demonstrate how their proposed projects will comply with our approach and expectation towards TR&I, identifying potential risks and the relevant controls you will put in place to help proportionately reduce these risks. See further guidance and information about TR&I , including where you can find additional support.

We are running this funding opportunity on the new UKRI Funding Service, so please ensure that your organisation is registered. You cannot apply on the Joint Electronic Submissions (Je-S) system. The project lead is responsible for completing the application process on the Funding Service, but we expect all team members and project partners to contribute to the application.

Only the lead research organisation can submit an application to UKRI. Select ‘Start application’ near the beginning of this Funding finder page: Confirm you are the project lead. Answer questions directly in the text boxes.

You can save your answers and come back to complete them or work offline and return to copy and paste your answers. If we need you to upload a document, follow the upload instructions in the Funding Service. All questions and assessment criteria are listed in the How to apply section on this Funding finder page.

Allow enough time to check your application in ‘read-only’ view before sending to your research office. Send the completed application to your research office for checking. They will return it to you if it needs editing.

Your research office will submit the completed and checked application to UKRI. Please be aware that research office and finance teams undertake checks on hosting arrangements and financial eligibility. The ultimate responsibility for ensuring compliance with all opportunity requirements lies with the applicant.

Where indicated, you can also demonstrate elements of your responses in visual form if relevant. When including images, you must: provide a descriptive caption or legend for each image immediately underneath it in the text box (this must be outside the image and counts towards your word limit). insert each new image onto a new line.

use files smaller than 5MB and in JPEG, JPG, JPE, JFI, JIF, JFIF, PNG, GIF, BMP or WEBP format. Images should only be used to convey important visual information that cannot easily be put into words.

The following are not permitted, and your application will be rejected if you include: sentences or paragraphs of text excessive quantities of images A few words are permitted where the image would lack clarity without the contextual words, such as a diagram, where text labels are required for an axis or graph column.

For more guidance on the Funding Service, see: how applicants use the Funding Service how research offices use the Funding Service how reviewers use the Funding Service References should be included within the word count of the appropriate question section. You should use your discretion when including references and prioritise those most pertinent to the application. Hyperlinks can be used in reference information.

When including references, you should consider how your references will be viewed and used by the assessors, ensuring that: references are easily identifiable by the assessors references are formatted as appropriate to your research persistent identifiers are used where possible General use of hyperlinks Applications should be self-contained. You should only use hyperlinks to link directly to reference information.

You must not include links to web resources to extend your application. Assessors are not required to access links to conduct assessment or recommend a funding decision. Generative artificial intelligence (AI) Use of generative AI tools to prepare funding applications is permitted, however, caution should be applied.

For more information see our policy on the use of generative AI in application and assessment . UKRI must receive your application by 20 October 2026 at 4:00pm UK time. You will not be able to apply after this time.

Make sure you are aware of and follow any internal institutional deadlines. Following the submission of your application to this funding opportunity, your application cannot be changed, and submitted applications will not be amended. If your application does not follow the guidance, it may be rejected.

UKRI will need to collect some personal information to manage your Funding Service account and the registration of your funding applications. We will handle personal data in line with UK data protection legislation and manage it securely. For more information, including how to exercise your rights, read our privacy notice .

If you or a core team member need to tell us something you wish to remain confidential, email ai. roboticsEPSRC@epsrc. ukri.

org Include in the subject line: [the funding opportunity title; sensitive information; your Funding Service application number].

Typical examples of confidential information include: individual is unavailable until a certain date (for example due to parental leave) additional information about eligibility to apply that would not be appropriately shared in the ‘Applicant and team capability’ section conflict of interest for UKRI to consider in reviewer or panel participant selection the application is an invited resubmission For information about how UKRI handles personal data, read UKRI’s privacy notice .

Institutional matched funding There is no requirement for matched funding from the institutions hosting the project lead, project co-leads or other staff employed on the application, beyond the standard 20% FEC. Expert reviewers and panels assessing UKRI funding applications must not consider levels of institutional matched funding as a factor on which to base recommendations.

Direct and in-kind contributions from third party project partners are encouraged. This policy does not remove the need for support from host organisations who must provide the necessary research environment and infrastructure for award-specific activities funded by UKRI. For example, research facilities, training and development of staff.

Engineering and Physical Sciences Research Council (EPSRC), as part of UKRI, will publish the outcomes of this funding opportunity at EPSRC Funding Applications Outcomes . If your application is successful, we will publish some personal information on the UKRI Gateway to Research . In plain English, provide a summary of your application.

We usually make this summary publicly available on external-facing websites, therefore do not include any confidential or sensitive information.

Make it suitable for a variety of readers, for example: the wider research community Guidance for writing a summary Clearly describe your proposed work in terms of: the challenge the project addresses potential applications and benefits List the key members of your team and assign them roles from the following: project co-lead (UK) (PcL) project co-lead (international) (PcL (I)) professional enabling staff research and innovation associate Only list one individual as project lead.

Joint leads may be named under ‘Applicant team and capability to deliver’. UKRI has introduced a new addition to the ‘Specialist’ role type. Public contributors such as people with lived experience can now be added to an application.

Find out more about UKRI’s core team roles in funding applications . Create a document that includes your responses to all criteria. The document should not be more than 3 sides of A4, single spaced in paper in 11-point Arial (or equivalent sans serif font) with margins of at least 2cm.

You may include images, graphs, tables. You can have an additional page for a diagrammatic workplan. For the file name, use the unique Funding Service number the system gives you when you create an application, followed by the words ‘Vision and Approach’.

Save this document as a single PDF file, no bigger than 8MB. Unless specifically requested, please do not include any sensitive personal data within the attachment. If the attachment does not meet these requirements, the application will be rejected.

The Funding Service will provide document upload details when you apply. What are you hoping to achieve with and how will you deliver your proposed work?

What the assessors are looking for in your response For the Vision, explain how your proposed work: is of excellent quality and importance within or beyond the field(s) or area(s) has the potential to advance current understanding, generates new knowledge, thinking or discovery within or beyond the field or area is timely given current trends, context and needs impacts world-leading research, society, the economy or the environment As this is a high-risk, high-reward opportunity, we will not be penalising risky ideas that can build a pipeline for impact.

Within this section we also expect you to: explain how the proposed work will advance the explainability of future AI systems For the Approach, explain how you have designed your work so that it: is effective and appropriate to achieve your objectives is feasible, and comprehensively identifies any risks to delivery and how they will be managed if applicable, uses a clear and transparent methodology if applicable, summarises the previous work and describes how this will be built upon and progressed will maximise translation of outputs into outcomes and impacts describes how your, and if applicable your team’s, research environment (in terms of the place, and relevance to the project) will contribute to the success of the work Within this section we also expect you to: provide a comprehensive project plan including milestones and timelines clearly describe both the framework and proposed methods and explain the reasons for their choice References may be included within this section.

Applicant and team capability to deliver Why are you the right individual or team to successfully deliver the proposed work?

What the assessors are looking for in your response Evidence of how you, and if relevant your team, have: the relevant experience (appropriate to career stage) to deliver the proposed work the right balance of skills and expertise to cover the proposed work the appropriate leadership and management skills to deliver the work and your approach to develop others contributed to developing a positive research environment and wider community You may demonstrate elements of your responses in visual form if relevant.

Further details are provided in the Funding Service. The word count for this section is 1,500 words: 1,000 words to be used for Résumé for Research and Innovation (R4RI) modules (including references) and, if necessary, a further 500 words for Additions.

Use the R4RI format to showcase the range of relevant skills you and, if relevant, your team (project and project co-leads, researchers, technicians, specialists, partners and so on) have and how this will help deliver the proposed work. You can include individuals’ specific achievements but only choose past contributions that best evidence their ability to deliver this work. Complete this section using the R4RI module headings listed.

Use each heading once and include a response for the whole team, see the UKRI guidance on R4RI .

You should consider how to balance your answer, and emphasise where appropriate the key skills each team member brings: contributions to the generation of new ideas, tools, methodologies, or knowledge the development of others and maintenance of effective working relationships contributions to the wider research and innovation community contributions to broader research or innovation users and audiences and towards wider societal benefit Provide any further details relevant to your application.

This section is optional and can be up to 500 words. You should not use it to describe additional skills, experiences, or outputs, but you can use it to describe any factors that provide context for the rest of your R4RI (for example, details of career breaks if you wish to disclose them). Complete this as a narrative.

Do not format it like a CV. In this section, please indicate: joint leadership arrangements, if relevant References may be included within this section. The roles in funding applications policy has descriptions of the different project roles.

Ethics and responsible research and innovation (RRI) What are the ethical and RRI considerations, implications and issues relating to the proposed work? If you do not think that the proposed work raises any ethical or RRI issues, explain why.

What the assessors are looking for in your response Demonstrate that you have identified and evaluated: the relevant ethical and RRI considerations, including both the research or topic area itself and the design and delivery of the project the wider implications of the proposed work, and how you will maximise the positive societal, environmental, and economic benefits arising from the project, while minimising unintended negative impacts, such as research misuse or accidental harm how you will manage these considerations throughout the lifecycle of the project If you are collecting or using data you should identify: any legal and ethical considerations of collecting, releasing and storing the data (including consent, confidentiality, anonymisation, security and other ethical considerations and, in particular, strategies to not preclude further reuse of data) formal information standards that your proposed work will comply with You may demonstrate elements of your responses in visual form if relevant.

Further details are provided in the Funding Service. Please refer to the UKRI position statement on funding ethical research and Responsible innovation for more information around our expectations on ethical and responsible research and innovation. Animal Involvement and ‘3Rs’ You must complete this section about how your proposed project will involve or impact animals.

If your project does not involve or impact animals, you must confirm this on the next page. What animals you are involving The severity of the procedures you are using Where the procedures will take place Welfare standards you aim to meet The relevance of your project to the development, validation or dissemination of the 3Rs You may also need to download, complete, and upload at least one set of additional questions.

You will be told how to do this towards the end of this section. To complete this section and check whether your project is in the scope of the questions, refer to the UKRI policy for research and innovation involving animals . UKRI policy relates to all animals in the Kingdom Animalia, including vertebrates and invertebrates.

Genetically modified organisms and biological risk You must complete this section if your project will include genetically modified organisms or genetic technologies. If you project does not involve genetically modified organisms or genetic technologies, you must confirm this on the next page.

the type of organism your project will involve and the procedures your project will include the intended use of the organism or genetic technology the genetic, biological and environmental risks of your project For more information, see UKRI’s guidance on genetic technologies . Human participation in health-related research You must complete this section about whether your project will include human participation.

If your project does not involve human participation, you must confirm this on the next page.

what type of human participation your project includes the project design for human participation the phase of the clinical trial whether the project will be in an NHS setting, if so how the project will be registered whether diversity and inclusion will be considered For more information, see UKRI’s guidance for human participants in research .

Resources and cost justification What will you need to deliver your proposed work and how much will it cost?

What the assessors are looking for in your response Justify the application’s more costly resources, in particular: significant travel for field work or collaboration (but not regular travel between collaborating organisations or to conferences) any equipment that will cost more than £25,000 any consumables beyond typical requirements, or that are required in exceptional quantities all facilities and infrastructure costs all resources that have been costed as ‘Exceptions’ You can request costs associated with reasonable adjustments where they increase as a direct result of working on the project.

For further information see Disability and accessibility support for UKRI applicants and grant holders . Where a funding limit is imposed on the opportunity, requested costs for reasonable adjustments may exceed the maximum funding amount. Assessors are not looking for detailed costs or a line-by-line breakdown of all project resources.

Overall, they want you to demonstrate how the resources you anticipate needing for your proposed work: are comprehensive, appropriate, and justified represent the optimal use of resources to achieve the intended outcomes maximise potential outcomes and impacts To what degree is your proposed work speculative and high-risk, with potential for high reward?

Please comment on how the proposal will do one or more of the following: revolutionise existing fields of research create new fields of research test new, unproven tools or techniques cause paradigm shifts in existing thought and knowledge support discovery that might lead to radically new and disruptive technologies otherwise produce a step change in the explainability of future AI systems Within this section we also expect you to detail why success in this project has the potential to generate significant rewards in the future and deliver against government ambitions set out in the AI Opportunities Action Plan and Industrial Strategy , noting that these rewards can be long-term and speculative.

What the assessors are looking for in your response Demonstrate that your proposed project and research is: exciting and potentially transformative suitably risky to potentially produce revolutionary research designed to generate long-term rewards aligned with government strategy References may be included within this section.

Please identify up to two members of the team involved who will participate in the distributed peer review process. This should include the project lead and a co-lead if appropriate. For each reviewer, you must provide: their email address.

Please ensure you use the same email address that this individual has used in the core team section of your application which UKRI research council their expertise broadly falls within (that is EPSRC, ESRC, AHRC, MRC, BBSRC, NERC or STFC) Note that your answer to this question will not be assessed. However, failure to provide at least one reviewer (maximum two) will result in your proposal being rejected.

Nominated reviewers do not need to have membership of a research council Peer Review College. The applicant team should only nominate individuals with the capacity, knowledge and confidence to provide constructive, comprehensive, quality reviews. Please check that nominated reviewers do not have existing commitments (such as a long period of annual leave) which would prevent them from acting as a reviewer throughout November 2026.

Reviewers will have three weeks to complete each assigned review. Reviewers will be assigned eight or four applications to review depending on whether the PL or the PL and PcL are contributing reviews. We understand that individual circumstances can be unpredictable, and that there may be emergencies in a reviewer’s personal life which may cause delays or prevent them