As part of the recently launched AI Campus, Apolitical is publishing a series of articles exploring AI adoption in government. These articles take on many forms, from op-eds written by academic experts to interviews with public sector leaders working on AI adoption themselves.

Apolitical’s Ula Rutkowska recently had the opportunity to speak with Paul Maltby, Director of AI Transformation in Government at Faculty.ai, about the state of AI transformation in government.

Before starting his role at Faculty.ai, Paul Maltby held numerous positions in the UK government, including the Chief Digital Officer at the Department for Levelling Up, Housing and Community and the Director of Data for the Government Digital Service.

Q: Could you share what you’re currently thinking about regarding AI transformation in government?

Faculty.ai has been operating for 10 years as an applied AI company, working with a mix of private sector clients and various government bodies in the UK. Our focus on AI encompasses a broad spectrum of activities, use cases, and tools, which have evolved over the years. Initially, we assisted analysts in government to use data science to assist in evidencing policy decisions for Ministers. However, our current focus is on using AI in services to make routine operational decisions. We are involved in several key areas.

💬 Let the author know what you think about AI adoption in government leaving a comment below

Firstly, we utilise AI and machine learning to identify outliers in large and complex datasets, which is crucial for detecting harmful activities such as fraud, terrorism or distributing dangerous content on social media platforms.

Secondly, we are heavily involved in predictive analytics. You can see this in our work to develop a three-week early warning system for Covid-19 in the UK. This system was designed to predict where resources would be needed, enabling proactive distribution of PPE equipment. Generally, our focus on predictive analytics aims to provide actionable insights that can guide real-world decisions, particularly in managing complex operational situations such as those found in emergency departments.

AI presents a significant opportunity to automate administrative tasks, marking a pivotal moment for the stability and improvement of public services.

The third area of focus is optimising case flow processes within government. We’re working on updating underlying case management systems and processes to triage cases and efficiently direct tasks to the appropriate individuals. There are quite a few benefits to optimising these processes, particularly in government settings where there are growing backlogs and there has been a reluctance to invest in system improvements due to budgetary constraints.

Lastly, we’re working on understanding how generative AI has the potential to impact daily lives within government operations.

Q: What would you say are the most significant opportunities for government?

In the face of backlogs, constrained budgets and limited prospects for increasing civil servant numbers in the UK, there is a pressing need to enhance the efficiency and productivity of government services. Despite significant advancements in digital technology within government, many operational processes and case-working aspects remain largely unchanged. AI presents a significant opportunity to automate administrative tasks, marking a pivotal moment for the stability and improvement of public services. This shift could allow for resource reallocation from resultant savings and, paradoxically, could help humanise public services by freeing up personnel for work requiring human empathy and interaction. Looking ahead, generative AI promises to enable new service models, business approaches and efficiencies, such as personalised tutoring on an unprecedented scale. While these possibilities are exciting, they also come with considerations and potential drawbacks that must be carefully evaluated, particularly regarding these technological advancements' implications and ethical considerations.

Q: Can you delve deeper into the considerations and potential drawbacks of AI adoption in government?

My perspective as someone committed to public service reform has always been to facilitate the introduction of AI carefully and safely.

A longstanding concern is the potential for significant bias and discrimination, as well as the "garbage in, garbage out" phenomenon, where the quality of input data directly affects the output. This issue persists as data tools become more sophisticated, raising concerns that enthusiasm for these tools might overshadow the necessity for careful and nuanced operation.

The risk of amplifying biases with machine learning, especially where training data is crucial, cannot be overstated. Bias is an inherent risk in all data, requiring constant vigilance. Concerns about artificial general intelligence (AGI) and superintelligence are valid and warrant attention, but they should not detract from addressing immediate issues like data bias and discrimination.

Another area of concern is the societal and industry-wide implications of AI use in the wider economy, especially when it comes to employment. Previous experience has shown that despite short term adverse effects on jobs, technological advancements don't lead to long-term job losses over the longer term. But the widespread adoption of large language models could certainly put pressure on certain jobs in coming years, and evidence of this is already emerging in some sectors. Governments should proactively consider strategies for managing potential industrial disruptions, including contemplating interventions and reskilling initiatives to mitigate technological threats effectively.

Q: How do you see the role of AI in government evolving over the course of the next five to ten years?

The trajectory of AI in government is evolving from a niche area for analysts and scientists to becoming an integral part of core operational services. The effectiveness of government operations can significantly improve when digital service teams are equipped with the skills to understand, commission and execute algorithmic work as a fundamental part of their service delivery—a practice not widely adopted yet.

Integrating AI into the mainstream of operational government digital services represents a major advancement. It is equally important for traditional policy sectors and the broader civil service to recognise AI's potential to disrupt and transform their work areas too. I envision a future where policy officials across the government proactively consider the implications of AI in their domains, contemplating both how their teams can utilise it and the broader impacts it is likely to have on the sectors of society and the economy that they oversee. This forward-thinking approach should prevent AI from causing unexpected disruptions by thoughtfully anticipating and managing its influence. Ideally, AI should not catch anyone by surprise but be seen as a tool for positive change, with its potential impacts and applications thoughtfully integrated into policy and operational planning.

Done reading? Make sure to share your own thoughts by leaving a comment below ⬇️

(Image credit: Pexels)