AI Strategy


Ensuring AI is embedded in all sectors and regions

Tom Rodden

Professor Tom Rodden is Chief Scientific Adviser to the Department for Digital Culture Media and Sport (DCMS) and a Professor of Computing at the University of Nottingham. Prior to joining DCMS he was Deputy Executive Chair of EPSRC where he was responsible for research strategy and acted as the UKRI lead in both AI and e-infrastructure, driving several large-scale initiatives that span multiple disciplines across UKRI. He is a fellow of the Royal Academy of Engineering, a Fellow of the ACM and the BCS.


  • Investment levels must be maintained if we are not to fall behind our competitors
  • AI needs to become a part of our everyday lives
  • Diversity and skills need to be promoted at all levels
  • Standards for AI will be critically important
  • International cooperation and coordination will be crucial in the coming years.

The National AI strategy highlights a number of important factors. One key item is the level of investment. Some £2.3 billion has already been invested in AI. That level of support has to continue. Another priority highlighted in the Integrated Review, is the drive to complete the transition to being a science superpower.

AI technology is developing fast, it is advancing quickly and we must not slow down or take the foot off the accelerator. Our academic and research base is making major advances which need to be transferred quickly into commercial success and value for the UK.

The National Strategy sets out how AI can move from being a distinct technology to something that is integral to society, a part of society, incorporating the values we have and working for us. Future AI will need to be a partnership between people and the technological elements.

The international position is also important. A large amount of resource has been spent on different aspects of AI across the world and the UK needs to be selective and strategic in its engagement.

Diversity and trust are located at the very heart of the Strategy. There are then three pillars: the first is investment in the long term needs of the ecosystem. This is critical: the AI ecosystem does not simply emerge, it needs to be grown. The second is fairness: a vital element. The pandemic has highlighted an exacerbated lack of fairness and diversity across the UK. We must ensure that benefits accrue to all and to every sector and region equally.

We need SMEs across all sectors to adopt new, modern AI tools: this will be a critical driver in terms of maturing the benefits of these advances.

Effective governance

Then, there is the need for effective governance. What are the underlying drivers for building the ecosystem and what are the mechanisms to make this happen? Diversity and skills need to be promoted at all levels. We do not want an AI that reflects a privileged, white middle-class view. Rather, the AI of our society should represent a diversity of views, a diversity of perspectives and a diversity of backgrounds.

Many of the issues facing AI have their roots in philosophy and epistemology, not in mathematics. We really need to think carefully about what we mean here by ‘AI’ and by ‘intelligence’. These are not simple concepts and they need to be properly analysed.

AI research and innovation need to be at the core of what we are attempting and need to be coordinated and focussed. Transdisciplinary and multidisciplinary approaches must work together and link to industry as well. Appropriate tools are needed to achieve this. ARCHER2, the UK National Supercomputing Service, has now been switched on: we need this level of computational power in the UK in order to achieve our goals. In addition, we need to work out the best way to use data, including Government datasets and the data of our citizens.

There are a large number of industries in the UK that are updating their data skills with AI technologies. But specifically we need SMEs across all sectors to adopt new, modern AI tools: this will be a critical driver in terms of maturing the benefits of these advances. Giving businesses access to the necessary level of skills and expertise is going to be important, particularly for companies who cannot create their own AI department, or hire their own AI expert.

Government has to identify those parts of its own activities where AI can make a difference. These include critical missions, such as net zero and the health of the nation, where AI can be used and showcased.

Finally, governance is a crucial issue. Standards for AI are going to be critically important. They will form and shape the nature of AI going forward. As a nation, we must embed our values within those standards while providing genuine leadership on a global AI standardisation landscape.

Defining AI is really challenging. How then do you standardise something that is notoriously hard to define and in so doing establish and promote good practice? Current initiatives are looking at standardisation from different perspectives, so bringing these all together will present a challenge. Building or using an AI system may involve a choice between six or seven different guidelines. Some degree of coordination will be essential. This is a role for Government, promoting best practice and approaches to governance.

The Office for AI is seeking to address all the challenges of AI across Government. It is working through its action plan, consulting with Government, industry and others. The Strategy is a good vehicle to promote coordination.

We need to think carefully about international cooperation. That will develop in various ways over the next five to six years, particularly in terms of the partnerships we build. International cooperation, for many people, involves nation states. Yet the emerging technical AI superpowers may not be nation states and we may have to revisit our approach to international cooperation and relationships as we go forward.

Building or using an AI system may involve a choice between six or seven different guidelines.`