Research communication
Most research never reaches the people who could use it. In 2024 Cobo won an Innovate UK grant to test whether AI could close that gap.
It does, but it can damage what travels across it. An AI tool will summarise a paper and attach a citation, and the citation gets read as proof the summary is right. It isn't: it proves a source exists but someone still has to check the interpretation.
We think a studio that specialises in communicating science should be able to say something about it that isn't just an opinion. This page is what we can say.
Where this started
This didn’t start with a client. In 2023 we found ourselves reading a great deal of published research for personal reasons, and ran into how hard it is to know what a paper actually says when you aren’t the specialist who wrote it. We wondered whether AI could help.
In 2024 we won an Innovate UK grant through the DCMS Create Growth Programme to find out, which included building a first working prototype with Matheson AI in Bristol.
That work became Peelback, now a separate company founded by the same two people who run Cobo. It went on to win its own Innovate UK award and took the question much further than we could have as a studio.
Supported by
What Peelback found
Peelback Ltd was incorporated in March 2025, directed by Emilie Fontaine and Barney Chandler. It uses AI to automatically turn published research into designed outputs: study snapshots, conference posters, microsites, slide decks, LinkedIn posts, and whatever else a finding needs to travel.
Custom infographics are built into every output from the first draft, editable block by block, so no other design software is needed.
Nothing is shared until the researcher has verified it. The AI drafts, the author decides.
Peelback’s Innovate UK award funded a six-month discovery project with the University of Exeter, run between November 2025 and March 2026. 152 people took part, from published researchers to policy, commissioning and industry professionals, patients and clinicians.
The project went in thinking about simplification and came out designing for translation: the same finding, communicated differently to different audiences, for different purposes.
The full report is published at peelback.ai. Peelback also wrote about the trust question behind it for the LSE Impact Blog: “When AI communicates research, who signs it off?”
Peelback is currently working with industry and academic partners to test the MVP, and is looking to run pilot studies with research-associated companies and institutions.
