What ARIA funds, and who can apply
ARIA funds research teams it calls Creators. Official applicant guidance, checked on 18 November 2026, says those teams "could be anyone, from individuals doing their own research and PhD students at a university, to startups and large organisations." Applications are welcomed from individuals, universities, research institutions, small, medium and large companies, charities and public sector research organisations [source].
There are two official types of funding [source]:
- Programmes, with a total budget of £10 million to £100 million. A programme director builds a portfolio of projects that are meant to work together. Calls are tied to that programme's goals, not to an open topic.
- Opportunity seeds, of up to £500,000 per project. These are smaller and less structured, for individual teams exploring paths that might later become a programme or stand on their own.
ARIA says it does not retain IP rights to the work it funds, generally does not require match funding, does not take equity itself, and places a cap on the equity that funded organisations can hold in spinouts commercialising ARIA-funded IP [source]. It also takes an active role in managing projects. Those are distinctive terms next to many UK grants; they are still not "free money with no conditions", because each call has a scope, a process and a due-diligence stage.
This sits beside Innovate UK and UKRI in UK grants for AI: Innovate UK, UKRI and ARIA explained. How to write a bid is covered in how to write a UK AI grant application.
Concept papers, full proposals and what happens after
Most, but not all, programme calls start with a concept paper. Opportunity seed calls do not [source]. The official purpose of a concept paper is speed: no more than three pages in the general guidance, so reviewers can judge feasibility and relevance before asking for a full proposal. ARIA will tell you whether it encourages or does not encourage a full proposal, but you can still submit a full proposal either way [source].
A full proposal is a detailed document covering the technical work, the team, administration, budgets, IP and conflicts of interest. Applications are scored against the criteria in the solicitation. Final selection also looks at the programme portfolio as a whole, including the diversity of applicants [source]. You may submit more than one proposal to a call if each is a distinct project; applying for a seed does not stop you applying to other ARIA calls [source].
If you are selected, the award is still subject to due diligence and negotiation. ARIA expects an initial call between the programme director and the lead researcher within ten working days of notification, and contract or grant signature no later than six weeks from the successful or unsuccessful notifications [source]. Organisation consent is not required at concept-paper stage, but it is required at full proposal if you want to do the work at that organisation [source].
Joint applications should name a lead applicant. ARIA funds the lead, which is then responsible for agreements with collaborators, who should be listed as subcontractors or sub-grantees [source]. ARIA staff, their family members, and entities in which they are a director or person of significant control, are ineligible [source].
Does the work have to be in the UK?
Unless a call says otherwise, ARIA's primary focus is on people based in the UK, and for the vast majority of applicants it requires the majority of the project work to be conducted in the UK: more than 50% of project costs and personnel time [source].
It can still award funding where the work will primarily take place outside the UK, if it believes that would boost the net impact of the programme. In that case you must set out UK plans or commitments during the project, and those plans become part of the negotiation if you are selected. If you intend to create a UK subsidiary, it does not have to exist before funding starts, but you need a credible plan to achieve that during the project [source].
Some individual calls are more open on location than the default rule. Read the solicitation, not only this summary. The Robot Locomotion programme page, for example, says applications are eligible from any type of organisation and from any location, with a UK benefit case expected, especially for the more commercially focused technical area [source]. The default "majority in the UK" rule still applies unless the call says otherwise [source].
Calls open as of 18 November 2026
These are the ARIA listings we currently track. Always confirm dates on the official programme page, because a solicitation can close or be clarified after we last checked it.
- ARIA Robot Dexterity: Benchmarks, Standards and Interoperability call. The programme is backed by £88.5 million. The open request for proposals is for benchmarks, standards, interfaces or other shared capabilities in dexterous robotics, with tactile-sense data standards a particular priority. Proposals are a PDF of no more than five pages, plus references, submitted through ARIA's portal. The official deadline is 1 December 2026, 13:00 GMT, with notification of selection for discussions on 9 December 2026. Clarification questions for the ARIA team should be submitted no later than 24 November 2026 [source].
- ARIA Robot Locomotion programme call (concept papers). The programme is backed by £60 million. Concept papers are optional but strongly encouraged: a maximum of five A4 pages for Technical Area 1 (integrated robot bodies) and three for Technical Area 2 (foundational technologies). The official concept-paper deadline is 4 December 2026, 14:00 GMT. ARIA aims to say whether it would encourage a full proposal by 14 December 2026. The full-proposal window opens on 15 December 2026, 12:00 GMT, and closes on 1 February 2027, 14:00 GMT. Applicants are notified on 15 March 2027. Clarification questions for concept papers should be in by 27 November 2026 [source]. Software, AI or autonomy research without a substantive physical advance is out of scope for this call [source].
- ARIA Safeguarded AI: Cybersecurity (Track 1) rolling funding. The programme budget on our listing is £59 million, with no amount stated per award. Applications are described as rolling and reviewed in quarterly batches. Our last check, on 1 November 2026, noted that the 31 October 2026 cut-off had passed, so new applications go to the next cycle, with dates then to be confirmed. Check the official programme page before you prepare a proposal [source].
Programme pages usually quote a programme-level budget, not a typical award size. The call documents, not the headline £ million figure, tell you what a single project can ask for.
What to do next
- Read the official applicant guidance, then the specific call, including what is out of scope. Many AI-only software ideas will not fit a hardware-led programme.
- If a concept paper is invited, write that first. It is short on purpose. Use the call's own page limit, which can differ from the general "no more than three pages" guidance.
- Estimate costs at the level the stage requires: a high-level table at concept-paper stage, and a detailed breakdown at full proposal, using ARIA's cost sheet [source].
- If you want help with the writing, draft the technical idea and the budget yourself first. Grant-writing help: when to hire a bid writer covers when that is worth it.
Open grant calls, including these, are on the grants list, soonest deadline first.
