Can Public Sector Administrative Burden Be Improved with AI? Research Offers Recommendations for Next Government

Artificial intelligence (AI) could save UK public sector workers 23 million hours of administrative work a week, new research suggests.

Research by Microsoft and Dr. Chris Brauer, Goldsmiths, University of London, found that AI could save more than four hours a week on administrative tasks per employee across all public sectors.

As reported, given that there are around 5.93 million people working in the UK public sector (as of December 2023), this equates to a saving of 23 million hours per week.

More than half of UK public sector workers say administrative burdens are affecting their efficiency and motivation, while a report by Alan Turing in March suggested AI could help automate around 84 per cent of repetitive transactions across 200 government services.

However, research by the National Audit Office (NAO) has found that while 70 per cent of government bodies are piloting and planning AI use cases, the technology is not yet widely used across government.

Government departments are required to develop AI implementation plans by June 2024. However, the NAO reported that as of autumn 2023, just over a third (37%) of 87 government departments had implemented AI, with each typically having one or two use cases.

Administrative burden impacting employee productivity and morale

New research based on surveys of more than 1,000 public sector employees reveals that managing information and data is the administrative task that takes up the most time in the public sector, with each employee spending an average of more than eight hours a week on it.

As noted, the volume of administrative work has a significant impact on staff productivity and morale.

• 45 per cent of public sector respondents say they are “drowning in unnecessary red tape” and that this heavy administrative workload is having a negative impact on their mental health and wellbeing.
• 55 percent of respondents believe that the amount of administrative work negatively affects their ability to perform their daily work duties, while 54 percent believe that it reduces their job satisfaction and motivation.
• Half of respondents also said that heavy administrative workloads negatively impact the quality of services provided (48 percent) and limit the time they can devote to clients or patients (49 percent).

As a result, 57 percent of all public sector workers say they would prefer to spend less time on administrative tasks, and more than half (52 percent) feel they are constantly playing catch-up due to the amount of administrative work they have to do.

“Rapid advances in artificial intelligence (AI) are marking a turning point for public sector organizations around the world,” said Dr. Brauer. “Generative AI and large language models (LLMs) have the transformative potential to change government operations and redefine the future of public service delivery. Governments cannot afford to stand idle while AI transforms the world around us.”

Key recommendations for the next government

The report also makes seven key recommendations to the next government that aim to expand and strengthen the deployment of AI across the public sector:

1. Create a national AI delivery hub: Create a centralised unit in Whitehall to focus on AI deployment across government departments. This hub would bring together expertise and resources from government, academia and industry to develop scalable AI solutions.
2. Declare “AI for All” Principles: Introduce a set of guiding principles to ensure that all public sector employees benefit from AI while protecting themselves from potential threats. This declaration would emphasize the need for AI literacy and continuous learning.
3. Implement a comprehensive skills strategy: Develop a holistic approach to skills development for public sector employees. This strategy will target public sector leaders, employees and society at large, supporting a culture of continuous learning and innovation.
4. Unlock the power of public sector data: Break down data silos and treat data as a strategic asset. This includes modernizing data infrastructure and promoting data sharing across government entities to enable seamless AI integration.
5. Rethink procurement processes: Update procurement mechanisms to prioritize AI technologies. An AI-First policy would ensure the procurement of cutting-edge technologies that drive transformational change while upholding ethical standards.
6. Accelerate AI adoption in local government: Support local governments in scaling successful public-private partnership models. This includes creating a network of “change agents” who can drive AI adoption and share best practices.
7. Maximising economic opportunity: Update the Treasury’s approach to strategic investment in AI technologies across the public sector. This includes introducing an AI Transformation Mandate for all departments to identify and implement AI opportunities.