AI Champions Frontier AI Phase 1: AI-Enabled Health and Life Sciences is aimed at UK SME projects that can prove technical feasibility in frontier AI for predictive healthcare, medicines discovery, medicines manufacturing optimisation, or genomics and multi-omics. This is a Phase 1 feasibility competition, not a deployment programme. Applicants must be UK-registered SMEs, apply alone, keep total project costs between £150,000 and £250,000, and deliver a 3 to 6 month project. Collaborations and subcontractors are not allowed in this round.
The timing is good. Government says AI is already being embedded in UK healthcare and science, with one-third of NHS chest X-rays, or 2.4 million scans, now AI-assisted, while the AI for Science Strategy says AI is reshaping life sciences across research, diagnostics, treatment and manufacturing.
This theme is not looking for a generic healthtech product with an AI layer added on top. It is looking for core AI and ML novelty that can deliver a step change in capability.
In practice, that means assessors will want to see that your advance comes from one or more of the areas Innovate UK names explicitly:
For Theme 1, your project must sit in one of these four priorities:
This theme is not looking for a generic healthtech product with an AI layer added on top. It is looking for core AI and ML novelty that can deliver a step change in capability.
In practice, that means assessors will want to see that your advance comes from one or more of the areas Innovate UK names explicitly:
For Theme 1, your project must sit in one of these four priorities:
A strong Theme 1 application usually has one tight technical story.
For example, it might focus on:
Where many applications go wrong
Most weak applications in this theme fail for one of five reasons.
Define the technical problem precisely: State the scientific or engineering gap in plain English
Name the uncertainty: Be explicit about the one thing you do not yet know, and why a competent professional could not resolve it easily at the outset.
Build the smallest meaningful proof: Do not over-scope. Phase 1 is short. Prove the architecture, subsystem or learning approach that matters most.
Benchmark hard: Use quantitative success criteria from day one.
Finish with a credible Phase 2 bridge: Innovate UK says Phase 1 projects must end with a technical white paper and a clear translation into Phase 2 objectives.
FAQs
