Brieflist

Efficient, Structured and Controllable AI Systems

This grant has closed. Similar open grants are listed below. Get the next one like it, or the free weekly digest:

UK registered SME businesses can apply for a share of up to £1.66 million to deliver experimental validation of prototype AI systems for novel reusable components of efficient, structured and controllable learning systems.

Read the official page on Find a grant ↗ Get new grants like this

Funder
Innovate UK, part of UK Research and Innovation (UKRI)
Where
🇬🇧 United Kingdom
Status
Closed 18 Nov 2026
Opened
12 Oct 2026
Funding
£50k–£150k per award (£1.7m in total)
Who can apply
Small businesses
Topics
Education & training Innovation & technology Business & trade

About this grant

Description

Innovate UK, part of UK Research and Innovation (UKRI), will invest up to £1.66 million on industrial research for novel reusable AI mechanisms. This is subject to a sufficient number of high quality applications being received.

The aim of this competition is to advance the development of novel reusable artificial intelligence (AI) mechanisms that will form the foundations for the next generation of AI and machine learning (ML) systems of the future.

The driving strategic imperative is that future capabilities beyond the current state of the art, require fundamentally different underlying architectures and learning systems to advance beyond current limitations. Resource adaptive learning mechanisms, intervention tested and counterfactual prediction, graph-native learning topologies and representation modelling for scalable primitive learning components are fundamental components for adaptive scalable learning systems.

We will fund UK SMEs to experimentally validate novel AI mechanisms that could underpin future generations of efficient, structured, adaptive and controllable AI systems.

For this competition, novel mechanistic AI means an AI or ML system in which the principal technical advance is attributable to one or more of:

a new learning or credit assignment mechanism

an adaptive or dynamically reconfigurable architecture

a learned causal or intervention grounded model

a graph-native or relational computational architecture

a compositional representation, skill or operator mechanism

a mechanism, model or architecture exerting measurable control over an agent

When an existing frontier model is used, applications must demonstrate that the proposed advance remains technically material above the base model.

The competition is intended primarily for:

research intensive AI startups

deep tech SMEs

university spinouts

pre-commercial companies developing proprietary AI technology

Technology readiness level (TRL)

Projects must start at TRL 1, 2 or 3 and we expect projects to be making an advance of 0.5 to 1 TRL during delivery.

Your project must deliver all of the following:

a prototype algorithm architecture or subsystem

an experimental and validation environment

results against named baselines or a hypothesis

ablation and negative control evidence

a plan for further development

At the end of the project, you will be required to submit a technical whitepaper summarising:

technical progress and outcomes: concise description of what was built and learned

evidence and validation: quantified results against pre-defined success criteria, test conditions and a clear statement of remaining technical risks and limitations

innovation defensibility plan

Subject to Business Case approvals, there might be a second phase of this opportunity. If successful, we will notify you of the conditions and availability of Phase 2. There is no funding currently allocated for Phase 2.

This is phase 1 of a potential 2 phase competition.

In applying to this competition, you are entering into a competitive process. This competition has a funding limit, so we may not be able to fund all the proposed projects. It may be the case that your project scores highly but we are still unable to fund it.

Our experience from similar competitions suggests that you could have 4% chance of success.

We consider a range of factors when determining whether to provide funding to applicants. This includes an assessment of prior conduct, such as any outstanding payments owed to Innovate UK or UKRI. Such factors may influence the funding decision, potentially resulting in a refusal of funding or an award subject to additional scrutiny.

We also reserve the right to adjust funding allocations for any of our competitions. This may be in response to changes in policy, portfolio funding considerations or broader government funding decisions.

This competition closes at 11am UK time on the deadline stated in this Innovate UK competition …

Objectives

Your proposal

The aim of this competition is to advance the development of novel reusable artificial intelligence (AI) mechanisms that will form the foundations for the next generation of AI and machine learning (ML) systems of the future.

Your project must drive development of a new core AI or ML technology by establishing experimental evidence for a new architecture, learning process, representation or control mechanism.

The technology developed must have the potential to underpin capabilities, products, platforms or services across multiple future applications or markets.

We are looking for applications with a clear route to defensibility, such as protectable IP, proprietary data advantage, specialist know-how or other credible barriers to entry, with a path to scale.

Your project is not required to develop a complete commercial product. You must demonstrate the technical feasibility of the core mechanism, architecture or critical subsystem being developed.

Projects must start at a Technology Readiness Level (TRL) of 1 to 3 at time of application and we expect projects to be able to make an advance by 0.5 to 1 TRL level during project delivery.

Your proposal must demonstrate that AI or ML innovation is the core technical contribution and principal source of any potential competitive advantage you are likely to generate if successful.

Your project must deliver:

experimental validation of your hypothesis

a clear validation methodology against predefined metrics

evidence of technical novelty and a technical asset

a scaling rationale for the technology and business model

a technical white paper

To be in scope of this competition you must sufficiently describe:

the technical novelty and hypothesis driven experimentation statement

what technical asset the project will create

how it could underpin future products, platforms or services

the justification of TRL classification

Portfolio approach

We want to fund a variety of projects across different technologies, markets, technological maturities, themes and research categories. We call this a portfolio approach.

Specific themes

Your proposal must fall within one or more of the following themes.

You must focus on at least one specific priority within each theme you select. If your application does not align with the theme and specific priority you select, it will not be sent for assessment.

Theme 1. Resource and data adaptive learning

Learning mechanisms that reduce dependence on large datasets, repeated full model training or unnecessary computation by adapting what information is acquired, retained or updated according to uncertainty, task requirements or expected information value.

You must focus on one of the following priorities:

active and probabilistic learning, including uncertainty aware selection of informative data, experiments, simulations or interactions

continual and selective learning mechanisms that enable new knowledge or capabilities to be acquired without repeated full model retraining

data efficient learning under limited or expensive observations, using probabilistic representations, structured priors or uncertainty to improve learning efficiency

Theme 2. Causal and intervention grounded models

Mechanisms that learn, discover or exploit causal and counterfactual structure to improve generalisation, prediction, intervention selection or decision making when underlying conditions or data generating processes change.

You must focus on one of the following priorities:

causal representation learning

interventional or counterfactual learning

causal discovery under partial observability

Theme 3. Learned modular computation and routing

Learning mechanisms that discover, select, compose or reuse internal computational components, enabling models to dynamically route computation and reuse learned functions across tasks, environments or novel combinations.

You must focus on one of the following priorities:

learned modular …

Dates

16 October 2026

Online briefing event: watch the recording

Briefing slides are available to download from Supporting Information.

20 January 2027

Applicants notified

1 April 2027

Project start by

Eligibility

Who can apply

Your project

Your project must:

have total costs of between £50,000 and £150,000

last between 1 and 3 months

start by 1 April 2027

end by 30 June 2027

Any organisation receiving funding must carry out its project work in the UK, intend to exploit the results in the UK, and spend most of the funding within the UK.

Projects must always start on the first of the month, even if this is a non-working day. You must not start your project until your Grant Offer Letter has been approved by Innovate UK. Any delays within Project Setup may mean we need to delay your project start date.

You must only include eligible project costs in your application. See our overview of eligible project costs.

Lead organisation

To work alone your organisation must be a UK registered micro, small or medium sized enterprise (SME).

More information on the different types of organisation can be found in our Funding rules.

Subcontractors

Subcontractors are not allowed in this competition.

Number of applications

An eligible organisation can only lead on one application. Any further applications by the same organisation as lead will be made ineligible.

Sanctions

This competition will not fund you, or provide any financial benefit to any individual or entities directly or indirectly involved with you, which would expose Innovate UK or any direct or indirect beneficiary of funding from Innovate UK to UK Sanctions. For example, through any procurement, commercial, business development or supply chain activity with any entity as lead, partner or subcontractor related to these countries, administrations and terrorist groups.

Use of animals in research and innovation

Innovate UK expects and supports the provision and safeguarding of welfare standards for animals used in research and innovation, according to best practice and up to date guidance.

Applicants must ensure that all of the proposed work within projects, both in the UK and internationally, will comply with the UKRI guidance on the use of animals in research and innovation.

Any projects selected for funding which involve animals will be asked to provide additional information on welfare and ethical considerations, as well as compliance with any relevant legislation as part of the project start-up process. This information will be reviewed before an award is made.

Previous applications

You cannot use a previously submitted application to apply for this competition.

We will not award you funding if …

Applying for grants? Grants Pro e-mails you every morning with the new grants and opened forecasts that match your saved searches, and adds deadline reminders in your calendar, a private RSS feed and CSV export. Try it free for 7 days, then $9/month, cancel any time. See what Pro does →

Applying? Read how to write a winning grant proposal.

Contains public sector information licensed under the Open Government Licence v3.0 (Find a grant). Last checked 18 Nov 2026. Always read the official page before applying: it has the full conditions, documents and how to apply.

Similar grants