Public Sector AI Adoption Index 2026
Closing the Gap Between Promise & Practice
A project by Public First for the Center for Data Innovation, sponsored by Google
February 2026
What we set out to do:
Governments are entering a critical phase in the adoption of AI. We are already beginning to see the ways that AI is contributing to everyday public sector work. The question is no longer whether to adopt AI, but how to do so both effectively and responsibly.
The opportunity, if governments get this right, is vast. Many leaders recognise this, and have set out ambitious AI strategies to improve public services and deliver better outcomes for citizens.
But what is written in government strategies does not automatically translate into real-world impact. Approaches across countries are varied, with uneven progress in turning these ambitious plans into everyday realities.
We set out to understand the gap between promise and practice. We wanted to explore how governments and public servants are using, and are empowered to use, AI in practice. This research examines how AI is experienced by public servants every day — the human side of the story.
These insights are brought together into a global index that goes beyond measuring adoption and looks to the conditions that enable effective use. We also set out practical steps governments can take to turn AI ambition into real-world impact for the communities they serve.
What we did:
We surveyed 3335 public servants across 10 countries:
What our Index measures - The 5 Indicators:
Our index is based on five key indicators: enthusiasm, education, enablement, empowerment, and embedding. They help us identify strengths and areas for improvement within each country.
The 5 indicators that make up the Index are mutually reinforcing: high enthusiasm without enablement stalls adoption; access tools without the right empowerment can create risks and underutilisation of AI technology; education without embedding limits the impact and potential that AI can bring.
Enthusiasm
Measures how positive public servants feel about AI, whether they say it is useful or enhances their work, and how often they report positive, motivating experiences.
Education
Examines whether public servants have access to AI training, whether they have confidence in their own ability to use AI, and whether skill gaps or lack of guidance are reported as barriers to use of AI.
Enablement
Evaluates the access public servants have to approved AI tools, technical support, and organisational investment. Such as whether workers use workplace tools rather than personal or shadow systems.
Empowerment
Assess the clarity of rules, leadership direction, and whether public servants feel confident their AI use is permitted, appropriate, and aligned with workplace policy.
Embedding
Captures public servants’ views of how far AI is built into everyday systems and processes. Including integration with existing systems, and what organisational support exists for scaling use.
Our Key Findings:
We are in a moment of rapid AI uptake in the public sector
There is clear enthusiasm from public servants, 74% now use AI, with most adoption happening in the past year.
Effective use depends on clear leadership and support
In countries with clear guidance and backing for AI use, 91% of public servants feel confident using AI, 82% are optimistic, and 79% say AI feels empowering — compared with far lower confidence where rules and support are unclear.
A challenge for governments to go further
Public servants want their governments to invest more in AI for the public sector – just 18% think that Government’s are currently using AI very effectively
Index data:
Our index finds significant differences in how AI is experienced by public servants. In some countries public servants are better primed to fully realise AI’s potential than others.
Higher scoring countries perform consistently well across the index. This makes them better positioned to pull ahead with an ambitious approach to transforming public services, while others will struggle to see beyond more basic uses unless they take immediate action to create a more conducive environment for AI use.
Results – How Countries Stack Up:
Advanced Adopters
Singapore, the Kingdom of Saudi Arabia, and India. Each stands out across enthusiasm, empowerment and education.
- Saudi Arabia’s clear, top-down national strategy frames AI as modernisation rather than disruption. Public servants report clear permission and leadership backing.
- Singapore shows a system-led approach: cautious but effective, with standardised tools, clear authorisation and strong embedding into everyday work.
- India’s adoption is driven by enthusiasm and momentum at scale, even as systems and integration lag.
Uneven Adopters
The UK, U.S.A, South Africa and Brazil. They show patchy progress, with some leadership in enthusiasm and education. AI is present but unevenly embedded, often constrained by infrastructure or unclear permissions.
Cautious Adopters
Germany, France and Japan, They remain cautious and risk-averse, with opportunities to make impactful progress across all indicators. There, public servants leveraging AI are limited to specialist projects rather than everyday workflows.
What we learned, and lessons for leaders:
There are also clear lessons for leaders from each of the individual indicators – what effective AI adoption looks like in practice and where it breaks down. Below we set out key learnings from the data, and what this means for all government leaders as they seek to drive adoption and bridge the gap between promise and practice.
Enthusiasm
Strong enthusiasm scores show that public servants are optimistic about AI in the public sector, see it as effective, empowering, and expect (and want) it to have a meaningful, positive impact on their job.
For example, in India, which scores high for enthusiasm (71/100), 83% of public sector workers are optimistic about the use of AI technology and 59% would like AI to dramatically change their work day to day. The Indian Government’s ‘AI for All’ strategy has sought to encourage AI uptake across the country, with free courses hosted on the Government’s website and a clear and consistent positive message about the societal benefits of AI tools. Such messages can help drive up enthusiasm and interest in AI.
Enthusiasm - what we learned:
Enthusiasm drives uptake, especially for higher-value activities.
In high-enthusiasm workplaces, 86% of public servants report AI use increasing over the past year, compared with just 27% where enthusiasm is low. High-enthusiasts also tend to use AI across a wider range of tasks, with a majority (67%) applying AI to complex activities.
Good systems fail without enthusiasm.
Even where AI is well embedded in systems, low enthusiasm means adoption stalls — 39% of public servants in environments with good embedding but low enthusiasm have never used an AI tool.
Enthusiasm creates an engaged workforce.
In high-enthusiasm workplaces, job satisfaction is significantly higher (67% vs 36%).
Steps leaders can take to increase enthusiasm:
Set clear permission from the top: Public servants are far more likely to use AI when leaders have a clear and consistent message on the benefits AI will bring to their organisation, and clearly say it is allowed and encouraged to help achieve public service goals. Simple, practical guidance and a digestible AI Acceptable Use Policy can remove uncertainty and signal that AI is a normal part of everyday work.
Reinforce permission with visible leadership: Regular messages from senior leaders, along with real examples of teams using AI to save time and improve services, help turn permission into confidence and momentum.
Reassure common concerns: Data management is a consistent barrier to AI across the countries. Working with trusted and approved providers, and creating clear rules on how to handle data will allow public sector leaders to reduce fear of mistakes, help reduce informal or “shadow” AI use and improve organisational resilience.