Q&A: How can the UK Government deliver in a new era of AI?

Where are some of the most popular use cases for AI being identified across government?

The Office for Artificial Intelligence’s guide to using AI highlights that a number of public sector organisations are already using AI for a range of tasks, from fraud detection to supporting citizens.

AI can be used throughout the grants management process and put a stop to public sector fraud. By leveraging a connected data platform, using AI, automation, and big data analytics, it is a way to modernise the full grants management lifecycle. Lower-level processes, such as updating applications, processing grants, and checking financial details can be automated to allow civil servants to spend their time on higher value and more complex tasks. Further, a fully automated grant process, which is powered by AI and big data analytics can be used to stop fraudulent activity in its tracks. Layering the AI with automation creates a system which is instantly responsive to the findings of the fraud process, and stops any false payments from being made.

AI also provides great operational efficiency, which lowers costs and enables faster and more personalised services for citizens. Departments like HMRC and the Department for Work and Pensions (DWP) use AI-powered robotic process automation (RPA) to handle repetitive tasks, such as processing claims and tax returns. AI can also be used to automate eligibility checks for applications, such as benefits and loans. This creates much faster approval processes, reducing backlog and providing better service for citizens. As well as this, leveraging AI-powered chatbots enables 24/7 support for citizens, which leads to a lower burden on call centres.

What are the barriers you’re seeing to adoption?

Fundamentally, there are institutional structures that are inhibiting AI adoption, not only that of funding. It may be the case that the mechanisms which have been put into place to reduce the risk of delivery failure and wasted money are the very reasons for the lack of adoption. With the majority of IT programmes being top-down approvals, rather than the bottom-up agile approaches which are encouraged by the Government Digital Service (GDS), it does not align with the fast-pace and flexible approach that AI-driven initiatives need.

In this, there must also be a cultural shift within government to create more openness and readiness to use AI and automation. AI will not replace humans; the AI tools will be working alongside us. Further, there is a skills gap when it comes to AI. So, greater training and education opportunities are needed, with the employment of AI specialists within government. Additionally, there is a belief that IT projects should be done exclusively in house, and this often leads to a disconnect between IT teams and those that will be using the system, creating another barrier to adoption. Here, a low code software development can transform how systems are built, and how those within the government interact with the software – it fosters a culture of collaboration.

There are also issues with data quality and availability of complete data sets, which impacts the training of the AI models. There are also risks of bias within the data and models, which is a concern. Across the government, we continue to use legacy systems which store data in silos, making data sharing and integration difficult. This fragmented approach to IT continues to be a barrier for AI adoption and scalability. 

There is an opportunity here for a coordinated AI strategy across government departments for efficient and streamlined processes and workflows. Projects within government should be driven by AI from the start to get them off the ground much faster, and using the right technology platforms to ensure that they are done in a coordinated fashion. 

Do you think there is a gap between central and local government when it comes to AI adoption?

Yes, there is a significant divide between central and local government driven by a difference in funding. Local governments are more likely to face budget cuts which halts their ability to invest in AI. As local governments compete for grants, it creates uncertainty within budgets and leaves little room for them to drive AI initiatives. Within this, at the local level, governments are usually focused on different priorities such as social care, housing, and waste management, which are not seen at the central government level.

Similar to central governments, however, there is also an AI skills gap at a local level. This impacts their readiness to invest in AI projects. Notwithstanding, at the central level, there is more scope and dedicated teams to work on AI projects. Additionally, both central and local governments use outdated IT systems, but more so at a local level. We have already seen that HMRC and DWP are using RPA and AI in their operations, but this is unlikely to happen for local governments. This is due to older IT infrastructure and the lack of monetary and time resources to support these processes.

How is the Government doing in terms of ushering in this era of AI in terms of programmes, regulations, funding, etc.?

Over the past few years, the UK Government has published a number of frameworks on safe use of AI. In 2021, the National AI Strategy set out a ten-year plan looking at how Britain can become a global AI superpower. Notably, this framework cites an increase in investing in AI research, development, and innovation. In a similar vein, recently, the UK AI Safety Institute signed a new agreement with Singapore which aims to deepen international AI safety collaboration. Ushering in this new era of AI will drive innovation, but with tight controls on AI to ensure safe and ethical use of the technology. 

Importantly, the EU AI Act which was put into place on August 1, 2024 is an indicator that AI is being taken more seriously, with strong guardrails being put in place for its use. Globally, different regions can choose their AI policies and regulation, but it is key for businesses across the globe to aim for the most comprehensive regulation which is standardised. This will put customer protection at the forefront, but will also allow there to be safe and ethical AI-driven innovation. While the King’s Speech did not include details of equivalent UK AI regulations, it is likely that the EU AI Act will be influential in whatever is drawn up by the UK government.

 The UK Government is also taking a proactive approach to invest in AI. This year, the Government unveiled an £800 million investment plan to leverage technology to reform public services, such as policing and healthcare, and boost the UK’s productivity. Education and training are also central parts of this new era of AI, with investment going towards these programmes.

What is Pegasystems’ role and how are you working with government customers?

Pega offers a low-risk and incremental approach to transformation within the public sector. It allows government organisations to seamlessly automate processes across channels and better engage with citizens. Through this, we are able to minimise backlogs and accelerate decision-making, as well as identifying fraud.

Additionally, at Pega, we believe that a more iterative approach to transformation should be taken, rather than a ‘big bang’. There are a number of challenges which the government must overcome. Some of which are unsupportable legacy systems, which is the main priority, a lack of digital skills within the government, the huge upfront cost of investment, long implementation timescales and internal conflicts between business and IT teams over what transformation is needed. Even though there are these issues, departments across the government should start small and adopt an agile low-code approach which allows business and IT teams to work together and does not require a high-level of technical skills. Leveraging this approach will allow teams to demonstrate value as the go – outcomes and savings will be delivered – before moving to the next phase of transformation.

Pega works with governments around the world including the US, Australia, Singapore, France, Germany, Spain and the Nordic countries. Within UK government Pega is already delivering real change for some of the largest government departments:

• With HMRC, Pega is supporting tax compliance and anti-fraud initiatives and also now delivering significant improvements in the way the department approaches citizen interaction.
• With Cabinet Office, the Pega platform is underpinning the new Civil Service Jobs application.
• Delivering the platform for the Home Office to register the approximately six million EU citizens who remained in the UK after we left the EU, under the EU Settlement Scheme.
• With DWP, tackling costly fraud and error in the UK benefits system through joined-up, investigative case management.
• Helping Defra track and control the recent outbreak of Avian Flu.
• Working with the Nuclear Decommissioning Authority to help safely decommission Sellafield.
• Supporting the Ministry of Defence to recruit the key skills it needs to predict and deal with a fast paced and changing environment.

Recent advances in generative AI, and AI more generally, have the potential to radically transform the delivery of public services, to enhance efficiency, and to improve citizen experiences. If the right action is taken now, it should not be long before the implementation of AI systems in government will evolve to a level where they can seamlessly analyse vast amounts of data, interpret complex patterns and generate insightful predictions, enabling governments to make data-driven decisions with unparalleled accuracy and facilitating the development of targeted policies and programmes that address societal needs more effectively.

AI can be used now to automate repetitive tasks and therefore enable government to allocate resources more efficiently. Administrative processes, such as issuing permits, processing licences and managing records can be streamlined and expedited through the integration of intelligent automation. This can significantly reduce bureaucracy, enhance transparency and improve productivity.