Public sector customer service satisfaction is just 25%, even as 91% of agencies explore Gen AI. To succeed, agencies need modern IT, quality data, and a phased approach focused on accessibility and citizen trust. AI can transform public sector customer service by enabling personalization and efficiency.

Public sector CX is ready for transformation

Meeting citizens needs efficiently, effectively, and with empathy through an integrated, personalized service: this is what AI could really look like in the public sector. Yet research by the Capgemini Research Institute reveals that just 25% of citizens are satisfied or very satisfied with the service they receive from government agencies.

Our new point of view explores why high-quality customer service matters in the sector and what makes it challenging to deliver. It describes the potential Gen AI tools and AI agents offer to elevate that service. And it sets out four steps government leaders in IT or operational delivery can take to start realizing those gains – once they have the right foundations in place.

“72% of respondents in public services cite rising customer expectations as a critical issue facing their customer service functions.”

Unleashing the value of customer service, Capgemini Research Institute, March 2025

Where public sector organizations are now

Perhaps surprisingly, our survey shows that 91% of public organizations are exploring, piloting, or have already integrated Gen AI into their customer service operations, compared to a cross-sector average of 86%. And more public sector leaders say they are already seeing benefits, including reduced operating costs and issue handling time.

But thanks to issues including fragmented systems, siloed structures, and a lack of the necessary tech foundations, satisfaction for both citizens and customer service employees remains low.

“In the UK, some people with a long-term disability or condition must interact with 40 services across nine organizations”

The state of digital government review, GOV.UK 2025

The foundations of AI-enabled customer service

Gen AI tools and AI agents are not magic wand governments can wave to automatically get great results. To be effective, they need to layer their unique capabilities on top of other key enablers of strong customer service:

  • A modern, integrated, and scalable IT infrastructure.
  • Good-quality data, with standard definitions and governance around data sharing
  • Unified CRM systems that provide AI and customer service employees with a full picture of the person they are serving
  • Frameworks and processes for keeping citizen data private, secure, and compliant with regulations across all service channels

Four steps to consider if you are thinking of adopting AI in customer service

  1. Identify the opportunities and build the case for change
    Begin by picturing what a great experience could look like for the people you serve. Move to a life events model where services are designed around moments that matter to citizens – like losing a job or starting a business – rather than departmental silos.
  2. Clarify the role of AI and humans
    Define how humans and AI should work together. Use Gen AI tools for transactional tasks while humans focus on empathetic, complex interactions. Train employees to use AI-powered tools safely and confidently.
  3. Start small and measure impact
    Deploy AI gradually and create new KPIs for hybrid teams, such as AI accuracy, citizen satisfaction, and accessibility. Begin with low-risk pilots like conversation summaries or automated email suggestions.
  4. Deliver new value for everyone with AI
    Ensure accessibility for all, including people with disabilities. Gen AI can simplify complex information, provide real-time translations, and tailor content to user needs for faster, more seamless experiences.

What to consider if you are thinking of adopting AI in customer service

Providing a seamless, personalized service will involve changing people, processes, and mindset. From identifying opportunities to improving accessibility, these steps will help governments start creating new value in customer service with AI.