What offers the greatest return on investment for the public sector when implementing Artificial Intelligence?

Published on: August 24, 2026

The digital transformation of public administrations has entered a decisive phase. Beyond the initial adoption of language models for basic queries, public bodies globally are deploying Artificial Intelligence (AI) solutions to maximize the impact of their operational mission, optimize the use of public funds, and respond more quickly to citizen demands.

Based on data from the global report conducted by Google Cloud and National Research Group surveying 251 public sector leaders in 19 countries, it is possible to analyze with mathematical precision in which areas generative AI and agentic AI are delivering a tangible and proven return on investment (ROI).

The new agentic era of public administration

The public sector is experiencing an accelerated transition from passive generative AI tools to AI agents, specialized language models with specific roles, context, and objectives to plan, reason, and perform tasks independently, with the ability to access data and interact with other systems or agents.

The adoption of this technology reflects a structural change in the public sector’s investment priorities:

Adoption in production: 55% of public sector leaders state that their organization already uses AI agents in production.
Deployment at scale: 42% of organizations have deployed more than 10 operational AI agents.
Future budget: 61% plan to allocate 50% or more of their future AI budget specifically to agentic projects.
Increased spending: 75% of public bodies indicate that their organization’s overall spending on generative AI has increased.

In which areas is AI most profitable?

Although the technology is applied in multiple administrative areas, the return on investment is not distributed equally across all departments. These are the areas that have had the highest return.

1. Citizen service and experience (41% ROI)

It is the application that offers the highest profitability and is the most widespread (59% adoption). Agents dedicated to citizen service can guide people through complex procedures, facilitate access to funding programs, and resolve regulatory queries clearly, directly, and uninterruptedly 24 hours a day. This substantially reduces the volume of manual work and wait times.

2. Productivity and research (35% ROI)

This section represents the most relevant strategic finding of the report: although productivity and research ranks fifth in adoption level (49%), it rises to second place in return on investment (35%). The ability of agents to synthesize technical documentation, analyze grant eligibility, and automate scientific research tasks generates an accelerated economic return.

3. Software development and engineering (34% ROI)

IT development in public administrations often absorbs a large part of the technological budget. The application of AI in engineering and infrastructure productivity allows for optimizing code, reducing production failures, and accelerating the deployment of new digital public services.

Although there is also data on the areas of: technical support, finance and accounting, innovation and product design, and security and cybersecurity operations. – See more.

Productivity analysis: Measurable results in the public workforce

The ROI of Artificial Intelligence in the public sector does not only come from external automation, but from the direct increase in workforce performance. The quantitative impact on work teams reveals outstanding data:

70% of public sector leaders report a direct improvement in their employees’ productivity thanks to generative AI.
46% of executives state that their workers’ productivity has at least doubled.
74% point out productivity improvements among information technology (IT) staff.
60% record direct productivity improvements in non-IT staff.
54% highlight greater speed in obtaining key data and information.
52% confirm an increase in the accuracy of executed tasks.

At what stage can public bodies be found?

To structure investments and maximize profitability, public institutions go through three maturity phases in AI adoption:

Level 1 (Simple tasks): Use of conversational assistants (chatbots), basic information retrieval, and generation of standard images or texts.
Level 2 (AI agent applications): Implementation of specialized agents to interact with citizens/clients or support specific creative and analytical tasks.
Level 3 (Multi-agent workflows): Advanced orchestration of multiple interconnected agents that plan, collaborate, and execute complex end-to-end processes without constant human intervention.

How to start scaling AI?

Despite strong profitability indicators, massive AI adoption in the public sector faces organizational barriers. Only 32% of public leaders claim to have comprehensive executive support from top management to accelerate these initiatives.

Moving towards higher-impact solutions requires overcoming institutional silos and ensuring firm leadership. The return on investment in the public sector is consolidated when AI is not treated as an isolated pilot project, but as a strategic tool oriented towards citizen service, operational research, and cybersecurity. Focusing budgets on specialized agentic agents is the fastest route to multiply the operational capacity of public institutions and guarantee maximum impact on their mission.

If you want to delve deeper into this data, discover real success stories from the public sector, and consult the complete checklist to calculate the ROI of your Artificial Intelligence projects, we invite you to read the complete eBook prepared by Google Cloud and Luce IT. You can download it for free at the following link: Download Google Cloud eBook: The ROI of AI in the public sector.

Frequently Asked Questions about AI in the public sector

What is an AI agent in the public sector?

An AI agent is a specialized language model that has defined roles, context, and objectives to plan, reason, and execute tasks autonomously. They have the ability to connect to databases and interact with other agents to complete complex processes.

What are the areas with the highest ROI when applying AI in public administration?

According to sector studies, the highest return on investment is obtained in citizen service and experience (41% confirmed ROI), productivity and research (35%), and software development (34%).

How much does public employee productivity increase with generative AI?

70% of public sector leaders observe productivity improvements, and 46% state that their employees’ productivity has doubled or more after implementing generative AI tools.