AI at pace, with accountability: UK SME supports delivery of the Strategic Defence Review.
SME helps power delivery of the UK’s Strategic Defence Review using AI
Oxford Dynamics, a UK small and medium-sized enterprise, has announced that its AI platform was used to support the delivery of the UK’s Strategic Defence Review (SDR) – a notable milestone the company says marks the first time AI has been applied to assist the development of a UK defence strategy.
Oxford Dynamics announced the involvement publicly on 3 June, shortly after the review itself was released on 2 June. According to the company’s Director and Co-Founder, Dr Edward Jackson, its AVIS (“A Very Intelligent System”) platform helped analysts process and organise a large body of evidence submitted to the review: around 8,000 responses from more than 1,700 contributors.
The system was applied to material gathered through the SDR call for evidence, which ran from 28 August to 30 September 2024. Oxford Dynamics said AVIS accelerated insight generation while allowing human reviewers to concentrate on higher-order reading and judgement.
Security and auditability were emphasised as core requirements for work of this sensitivity. Oxford Dynamics stated that AVIS ran in a secure, air-gapped environment and produced outputs designed to be traceable back to original source material, enabling conclusions to be checked and verified. The company also highlighted AVIS’s ability to handle a wide range of input types, spanning structured documents and free text through to video, images, and even sensor-derived data.
Co-Founder Shefali Sharma said the engagement was enabled through Defence Digital’s existing enterprise agreements, aligning with the Ministry of Defence’s Digital and Data Strategy. Oxford Dynamics secured an MoD “enterprise lite” agreement in September 2024 – described as a mechanism intended to make it easier for trusted UK technology SMEs to engage with the defence ecosystem.
Collaboration with a global prime
To deliver the SDR support, Oxford Dynamics worked closely with Palantir, which provided data integration and infrastructure capabilities that helped underpin the deployment. With that foundation in place, Oxford Dynamics said AVIS could deliver explainable summarisation, clustering, and search across large volumes of unstructured material. Dr Edward Jackson characterised the pairing as an example of a UK SME and a global prime operating in a complementary way to support national objectives.
Oxford Dynamics has also collaborated with several parts of the MoD ecosystem in recent years, including Commercial X, the Defence Artificial Intelligence Centre (DAIC), the Defence Science and Technology Laboratory (Dstl), and the Defence and Security Accelerator (DASA).
Crossing the “valley of death”
Shefali Sharma argued that early engagement and practical feedback from defence stakeholders were instrumental in refining AVIS from an R&D concept into something deployable. She credited a mix of grants, structured feedback, and technical interaction, particularly through DASA, DAIC, and Dstl, with helping the company traverse the familiar gap between promising innovation and operational adoption.
Both founders welcomed elements of the SDR that emphasise UK-built capability and faster innovation pathways, saying the use of a UK deep-tech start-up in a strategic, high-profile review signals the direction of travel for defence adoption of new technology.
