Written evidence submitted by the Ministry of Defence.

 

I welcome the Defence Sub-Committee’s inquiry “Developing Artificial Intelligence capacity and expertise in UK Defence”. AI technologies have enormous potential to transform every aspect of Defence activities, from the back office to the front line. The rapid, safe and responsible development and adoption of these game-changing technologies is therefore a key strategic priority for the MOD. As you will be aware, we set out our overall goals, vision and approach through the Defence AI Strategy (2022), highlighting the critical importance of developing strong collaborative relationships with partners across the UK’s AI ecosystem. I am pleased to submit the following written evidence in response to the Sub-Committee’s questions.

 

 

How clearly has the Ministry of Defence set out its priorities for the kind of
AI capacity and expertise it believes the UK defence sector should have, what priorities has it identified, and are these deliverable?

 

  1. The Ministry of Defence (MOD) works closely with partners across UK industry and academia to ensure that we have access to the cutting-edge skills, capabilities and systemic capacity necessary to modernise and sustain the UK armed forces. In particular, the UK’s thriving technology sector is a national asset, with world leading capabilities in many areas that are relevant for Defence. Through the Defence AI Strategy (2022)[1] we set out our intent to forge a more dynamic and integrated partnership with this sector, seeking to address barriers to frictionless collaboration, incentivise engagement and co-creation, and broaden our reach into sectors of the

 

economy that may not have previously considered working with the MOD. More generally, as set out in the Defence Command Paper Refresh (DCPR, 2023),[2] we are working to ensure that the UK’s defence industry is geared and incentivised to accelerate the development of new technologies with potential military applications, while also exploring ways to increase productivity and supply chain modernisation through investment in digitisation, automation and skills.

 

  1. We recognise that our partners in industry and academia require clear demand signals to plan against and ensure that their businesses and programmes are effectively structured to support Defence requirements and priority national objectives. Given the breadth of Defence business and the complexity of our operating domains, it is necessary to communicate our requirements, intent, and expectations in a number of different ways.

 

  1. At a strategic pan-Defence level, the MOD signals its overall ambition and intent for the adoption of new technologies through publications such as the DCPR and the Integrated Operating Concept (2022).[3] R&D priorities for AI and other technologies are set out through the Defence Technology Framework (2019),[4] Areas of Research Interest (annual),[5] the MOD S&T Strategy (2020)[6] and, of course, the Defence AI Strategy. We engage directly with partners in industry and academia through established channels like the Defence Suppliers Forum and are establishing a Defence and Security AI Ecosystem through the Defence AI Centre (DAIC). We also communicate more tactical requirements through specific channels, such as Hackathons sponsored by the DAIC or regular themed Innovation Calls through the Defence and Security Accelerator (DASA). We will continue to build on these engagement pathways, working with government partners and – where appropriate – international partners, to communicate a scaled demand signal to encourage civil sector investment in defence-relevant AI R&D.

 

  1. Defence’s overall objectives and ‘Priority Outcomes through AI’ were set out in the Defence AI Strategy. These outcomes were developed in consultation with industry stakeholders to provide broad guidance, noting the extremely broad range of potential use cases for AI within Defence. They are:

 

 

 

 

 

  1. These high-level priorities are fleshed out through the Defence AI Playbook that was recently launched at the DAIC Connect industry event. The Playbook provides an overview of the areas in which AI could be applied within Defence, dividing the problem space into six broad categories for AI capacity:

 

 

 

 

 

 

 

  1. In addition to the overview provided by the Playbook, the DAIC also plays an important facilitating role, holding regular engagement events which bring together MOD teams and a wide variety of Industry Stakeholders. As examples:

 

 

 

 

  1. Individual Front Line Commands (FLCs), Top Level Budget (TLB) organisations and Enabling Organisations are responsible for developing and owning their own plans for AI-enabled capability and services.[7] These plans align with the overall ambition set out in the ‘Priority Outcomes for AI in Defence’ set out above and are maturing rapidly as Defence AI Strategy implementation proceeds.

 

  1. Finally, the MOD is preparing a Joint Service Publication (JSP936) on ‘dependable AI’ which will set out the overall policy and procedural framework that Defence AI programmes (both internal developments and those developed by suppliers) will need to work within. This will include guidance on legal and ethical considerations, risk frameworks and quality & assurance arrangements.

 

 

What strengths and expertise does UK industry currently have in the field of Artificial Intelligence with defence applications?

 

  1. As noted in the National AI Strategy,[8] the UK is recognised as a thriving, world-class centre for tech start-ups, research, innovation, and digital industries, and as a top tier global power in AI. Areas of particular strength for UK industry include fintech, biotech, digital marketing & advertising, and security. The UK’s position as a global AI leader is further strengthened by our respected legal system, well-developed regulatory approaches, and influential Government bodies such as the Centre for Data Ethics and Innovation.

 

  1. Most new and emerging technologies like AI are inherently dual use; Defence exploits the same fundamental technologies that could be developed for civilian applications. Defence applications for AI stretch from the corporate or business space to the frontline, enhancing capabilities right across Defence. The broad range of potential applications include helping enhance the speed and efficiency of business processes and support functions; increasing the quality of decision-making and tempo of operations; improving the security and resilience of interconnected networks; enhancing the mass, persistence, reach, and effectiveness of our military forces; and protecting our people from harm by automating ‘dull, dirty, and dangerous’ tasks. The UK’s AI sector is well placed to support the development of game-changing solutions and capabilities across these use cases.

 

 

 

  1. The strengths and expertise of UK industry were recently demonstrated by the Defence and Security Accelerator (DASA) Innovation Focus Area, which attracted proposals from a range of organisations, including small and medium sized enterprises. Funded projects included proposals to automate change analysis from synthetic aperture RADAR measurements, capture and exploit data from operational radio traffic, derive economic insight using AI, and use Large Language Models to translate outputs from complex computer models of warfare into a simple narrative.

 

  1. The UK’s higher education system strengthens industry by nurturing a diverse skills-base, attracting elite global talent and spinning out countless innovative start-ups, complemented by a network of centres of excellence, innovation catapults and business accelerators, including the Alan Turing Institute, the Digital Catapult, Hartree Centre, Centres of Doctoral Training and Digital Research Infrastructures. Expertise on issues like AI ethics and safety is provided by organisations such as the Ada Lovelace Institute, Cambridge University’s Centre for the Study of Existential Risk and Leverhulme Centre for the Future of Intelligence, the Centre for Long-Term Resilience and Oxford University’s Centre for the Governance of AI.

 

 

How can the UK Government best develop capacity and expertise within domestic industry in sectors such as engineering and software to support the development and delivery of Artificial Intelligence applications in defence?

 

  1. In March 2023, the UK Science and Technology Framework identified AI as one of five strategic technologies that will be critical to the nation’s future security, prosperity and global influence (along with Quantum, Engineering Biology, Future Telecommunications, and Semiconductors).[9] The Department for Science and Technology (DSIT) is leading cross-cutting initiatives to implement the Framework and enhance domestic abilities to develop, incubate and exploit these technologies. This includes ambitious national strategies to catalyse the ecosystem and set clear demand signals for innovation, leveraging government’s buying power to support economic growth, and crucial interventions to strengthen the UK's skills base. This national effort will also help to ensure that technology issues are treated coherently, conscious of synergies and interdependencies at all levels of the technology stack and across UK technology sectors.

 

  1. The MOD closely supports these initiatives, working with partners across government to ensure that Defence interests and requirements are understood and that the Department can access the talent, innovations and industrial capacity that are critical to the development of future military capabilities. This is particularly important for challenges that can only be properly addressed at a national level by exercising all the levers available across government, such as critical skills shortages. However, there are also some areas where Defence’s unique requirements and operating context may necessitate slightly divergent approaches. For example, we strongly support the proposal set out in DSIT’s AI Regulation White Paper (August 2023)[10] for a pro-innovation, sector-specific approach to AI regulation that encourages consistency while recognising that a one size-fits-all regulatory framework could unintentionally frustrate or constrain R&D in some sectors.

 

  1. As highlighted above, the MOD’s overall approach and objectives for AI were set out through the Defence AI Strategy and the DCPR. As part of a wider programme to develop a digital response to the  Defence and Security Industrial Strategy (2021)[11] for the Defence Technology (‘DefTech‘) sector, we are developing more sophisticated and strategic relationships across the UK technology sector; driving research, supporting skills and professional development, promoting innovation, and encouraging joint investment to develop and sustain the industrial capability Defence needs, including through export and international collaboration opportunities. In respect of AI, much of this work is driven through the Defence S&T Portfolio or through the Defence AI Centre.  As examples:

 

 

 

 

 

 

governance structures, what the barriers might be and how we can address them. This builds on earlier work in partnership with Team Defence Information, which consulted 70 industry experts over several months to provide an industry perspective on AI production models and associated Assurance regimes. Analysis from these and further industry engagements will inform our Dependable AI Joint Services Publication later in the year.

 

 

 

What can the Government do to help embed UK AI companies in defence supply chains, both domestically and internationally?

 

  1. Consultations with partners across the UK AI sector have highlighted the following key factors to incentivising engagement and co-creation with UK AI companies:

 

 

 

 

Backbone will deliver data-driven, interconnected digital systems that integrate across domains, partners, allies and suppliers. Effective management, coherence and standardisation of data pan- Defence is central to the success of the Digital Backbone and a critical enabler of Multi-Domain Integration. 

 

 

 

 

 

 

 

 

How can the UK Government ensure that it champions the UK AI sector in the context of Pillar 2 of the AUKUS Partnership?

 

  1. AUKUS Pillar II provides the UK Government an opportunity to enable international partners to access and purchase world-leading capabilities within the UK's AI sector. Through joint development and experimentation with AUKUS partners, the MOD has demonstrated the strength of UK industry capabilities to our trilateral partners. Under AUKUS Pillar II, the UK Government, in collaboration with our trilateral partners, has created new channels through which we can champion the UK AI sector.

 

  1. The UK is collaborating with its AUKUS partners to deliver AI algorithms and machine learning to enhance force protection, precision targeting, and intelligence, surveillance, and reconnaissance. This effort builds on joint work demonstrated in the UK in April 2023 and in South Australia in October 2023. The AUKUS nations aim to integrate Resilient and Autonomous Artificial Intelligence Technologies (RAAIT) into national programs in 2024, to pursue the rapid adoption of these technologies across land and maritime domains. This is an ambitious trilateral programme of work which clearly demonstrates the UK’s commitment to championing the AI sector. In 2024, the MOD will convene the AUKUS Industry Forum, which will bring together trilateral government and industry representatives to help inform policy, technical and commercial frameworks.

 

  1. There is already significant UK industry contribution to AUKUS AI initiatives, through both the AI & Autonomy and the Undersea Warfare Working Groups, including support to trials with AUKUS partners. The AI & Autonomy Working Group has delivered multiple AUKUS trials with the support of UK SMEs capabilities. By testing UK industry offerings alongside those of our AUKUS partners, the MOD has been able to demonstrate some of the unique strengths of the UK AI sector. The Undersea Warfare Working Group has worked in partnership with industry across all three nations to develop algorithms that have recently been used to process data on US Maritime Patrol Aircraft for the first time.

 

  1. As AUKUS Pillar II continues to develop, the MOD and its partners will be seeking greater engagement from industry to support AI and other advanced capabilities. The Defence Ministers Meeting held on 1 December 2023 endorsed a range of announcements and initiatives which seek to continue our commitment to streamlining defence trade controls and information-sharing while minimising policy and financial barriers across public and private sectors. The meeting announced the creation of new mechanisms like the AUKUS Advanced Capabilities Industry Forum and AUKUS Defence Investors Network, which will facilitate cooperation between government and UK industry in crucial sectors like AI. The announcement of the Defence Trade and Industrial Base Collaboration initiative will further increase private-sector cooperation across the AUKUS partners in emerging technologies and strengthen resiliency across our supply chains.

 

  1. We are backing up this intent by working with industry to promote and facilitate AUKUS opportunities. To date there has been significant outreach from the UK Government with national industry, including through a series of classified briefing sessions providing direct access to government decision makers and delivery teams across the priority areas of engagement under Pillar II. This year, an AI and Autonomy classified brief was delivered to UK suppliers through the Capability Management International & Innovation (CMI&I). This built on the collaboration the AUKUS AI & Autonomy Working Group had already conducted with UK industry in support of the inaugural AUKUS AI and autonomy trial that was supported by two UK industry partners. It also provided an opportunity for two-way communication on capability priorities and readiness of the sector to respond. These sessions have been incredibly valuable, covering a range of issues from sovereignty and IP protection to production capacity and cooperation. UK Defence Solutions Centre are supporting the UK MOD on our sustained effort to build awareness across key sectors.

 

I am grateful to the Committee for its interest in this important subject. It is essential that Defence is able to work seamlessly with the brightest and the best minds across the UK AI sector to realise our ambitions to modernise and transform the UK Armed Forces at pace and scale. I look forward to reading the conclusions and recommendations from the Committee’s Inquiry, though please do let me know if you require more information about any of the themes that I have set out in this letter.

 

I look forward to hearing from you as the inquiry progresses.

 

 

 

2nd February 2024

[1] Defence Artificial Intelligence Strategy - GOV.UK (www.gov.uk)

[2] Defence Command Paper 2023: Defence’s response to a more contested and volatile world - GOV.UK (www.gov.uk)

[3] Integrated Operating Concept - GOV.UK (www.gov.uk)

[4] Defence Technology Framework - GOV.UK (www.gov.uk)

[5] Areas of research interest - GOV.UK (www.gov.uk)

[6] MOD Science and Technology Strategy 2020 - GOV.UK (www.gov.uk)

[7] Examples of such plans include: British Army Approach to Artificial Intelligence, Royal Navy Artificial Intelligence Adoption Roadmap

[8] National AI Strategy - GOV.UK (www.gov.uk)

[9] UK Science and Technology Framework - GOV.UK (www.gov.uk)

[10] A pro-innovation approach to AI regulation - GOV.UK (www.gov.uk)

[11] Defence and Security Industrial Strategy - GOV.UK (www.gov.uk)

[12] National Data Strategy - GOV.UK (www.gov.uk)

[13] Digital Strategy for Defence - GOV.UK (www.gov.uk)