Whilst AI is arguably one of the most exciting emerging technologies with the potential to improve services, detect risks, and connect and harness data for improved policy insights, there remains considerable caution over its deployment in the public sector. Concerns include skills to manage AI to collect, analyse and interpret data responsibly, algorithmic bias, potential costs of responding to the insights generated, and being able to take a more nuanced view of return on investment in making a business case for deployment.
In the UK, the issue of algorithmic bias and transparency is being addressed by the Algorithmic Transparency Recording Standard (ATRS), The standard offers a clear and accessible way for public sector organisations to provide information about the algorithmic tools they use in decision-making processes that affect members of the public.
In 2023, applications of AI will be concentrated in areas to drive simple automation and intelligent signposting to services, to using chatbots and linking to more effective application of robotic process automation (RPA).
We expect to see a growing interest amongst digital leaders in learning about and developing innovative AI applications to tackle deep seated and complex problems at a local, place-based level. In doing so, they will need to address some of the problematic areas of AI in order to win public trust, such as transparency in how algorithms work, unintended bias, vulnerability to new cyber risks and wider ethical considerations.
Two particular areas, social care support and traffic management in cities, are emerging as AI priorities for 2023. Cities in the UK are deploying traffic management systems that go beyond tracking congestion and controlling pollution, actively using AI sensors to change travel habits, support infrastructure management, improve road safety and journey planning, and predict incidents or where interventions are required. For example, 25 London boroughs are now using the VivaCity Sensor platform for these purposes.
(RPA) is becoming more common in public services, enabling a modular approach to process automation in parts of more complex systems. The main driver in 2023 is not just automation, but to create more adaptability by tuning or replacing RPA code when required. Some are even considering the application of AI to handle unstructured information in RPA.
‘Generative AI’ will begin to appear as a practical option in 2023, moving AI beyond intelligent signposting, automation and diagnostics, into the realms of policy planning, with scenario visualisation and simulation, including deployment of ‘Digital Twins’.
It will still take some time before more complex AI applications are trusted to link services and data sources to address complex citizen needs that cross different public services or target behavioural change. This is gradually becoming possible, and some are already looking at the opportunities and the inherent risks.
Case studies
New Zealand – AI to handle official information requests
“Artificial Intelligence is being used to trial handling of Official Information Act requests from the public. The ability to scan enormous amounts of unstructured data and return the relevant information will ensure local government can meet the statutory requirements of the Official Information Act without the time-consuming effort previously required.”
– Mike Manson, Chief Executive Officer, ALGIM, Aotearoa New Zealand
House of Commons – AI transparency and risks
MPs to examine regulating AI in new inquiry (commitees.parliment.uk)
The UK House of Commons Science and Technology Committee is examining the potential impact of artificial intelligence and the risks associated with it, with a white paper expected later this year. This will look at the “lack of transparency in how AI is applied and how automated decisions can be challenged”.
“AI is already transforming almost every area of research and business. It has extraordinary potential but there are concerns about how the existing regulatory system is suited to a world of AI.”
– Rt Hon Greg Clark MP, Chair of Science and Technology Committee, UK government
Leeds City Council – better outcomes with AI and RPA
Leeds City Council’s new digital strategy addresses the way in which artificial intelligence and robotic process automation (RPA) can improve data analytics, efficiency, and service outcomes.
“This digital strategy supports our key priorities focusing on improving the health and wellbeing of everyone, achieving our carbon zero ambitions, and ensuring we all benefit from the city’s growth and prosperity. No single organisation can achieve this alone; it must be a joint effort.”
– Councillor Debra Coupar, Deputy Leader, Leeds City Council, England
Auckland Council – RPA for building consents
“Auckland Council in New Zealand is using RPA (Robotic Process Automation) to improve the building consent process, taking an area of major concern for councils and turning it into a well-tuned application.”
– Mike Manson, Chief Executive Officer, ALGIM, Aotearoa New Zealand.