AI can deliver success for the public sector, but only if people remain at its centre

For years, digital transformation in government has focused on services, systems and technology. Yet one of the biggest barriers to better citizen experience is simpler and far more familiar. It’s content.

Content sits at the heart of almost every citizen interaction. Whether it’s service guidance, public communications, policy or internal knowledge – information is central to the public sector. Its quality shapes experience long before someone speaks to a caseworker or frontline adviser.

With decades of policies and legislation in place, as well as guidance and communications, public sector organisations are often made up of vast content estates. And they’re not always proactively managed as this can be very expensive and difficult to achieve.

And this brings fresh challenges for the public sector: getting content AI-ready so machines can easily find, interpret and use the right information at the right time.

Organisations simply can’t layer AI interfaces over content that’s out of date, irrelevant or even inaccurate. To do so is more than problematic – it can erode trust and break citizen confidence.  Without people-led content transformation, it can be hard for leaders to implement the technology as they’d like. 

Creating public sector content at scale

AI has shown that even for the public sector, it can provide valuable efficiencies. It can certainly identify duplication, analyse legislation, streamline governance and support content creation at scale.

For example, in Deloitte US, we’re there are good examples of where opportunities exist for government. For the State of Ohio, years of regulation had created a vast body of content that included outdated and duplicative requirements.

Deloitte developed RegExplorer™, an AI solution that enabled the state to examine thousands of regulations and millions of words to identify material that was duplicative, obsolete or redundant.

Over five years, Ohio reduced its regulatory code by a third and recovered around 58,000 hours of state employee time. Taxpayers are expected to save $44 million.However, for interpretation and context, content specialists, policy experts and service teams must remain responsible for citizen outcomes.

Why human-led content matters

As we heard in Deloitte’s State of the State 2026 research, citizens remain cautious about automated public services. They don’t want technology to remove the human element from government. It isn’t resistance to innovation. It’s a mandate to deploy AI intelligently: improving services while preserving its user-centred focus. 

Simply put, humans must ensure that decisions are being taken to enhance the citizen experience.

Done well, AI and humans working together creates both efficiency and trust: faster, more consistent content where risk is low, and real human involvement where decisions are complex, sensitive or high-stakes.

And this human input must be taken seriously, with clear quality controls in place.

People are needed to decide:

• what information actually means, particularly where policy or guidance is complex.
• which source is authoritative when information overlaps or conflicts.
• how citizens think about a service, rather than simply reflecting an organisation's internal structure.
• which terminology people understand, including people with different literacy, accessibility and language needs.
• what should be prioritised, combined, rewritten or removed which comes with some risk when dealing with legacy or historic documents at scale.
• where context and nuance matter, especially since the world and everything in it is evolving at a fast pace. Language is quickly outmoded, sentiments shift and humans need to be ahead of these changes.
• when information is sufficiently sensitive or consequential that a human needs to be involved, particularly where citizens’ welfare is at stake. 

The next challenge is transformation, not technology

Organisations should consider what their plan for a robust content strategy will be and how they can govern content to ensure it stays organised and machine ready.

This might include appointing ‘content champions’ within your organisation to ensure content transformation moves up the AI agenda and fast. Content can then be reviewed to ensure it is clearer and more dependable. Ownership should be reviewed also, as well as inconsistent standards and ways of working designed for a pre-AI era

Future AI value will come from transforming end-to-end processes, not adding isolated tools. Investing in technology is one thing, but the next phase requires the same attention to how content is created, governed and managed. Getting information in order is vital.

That is the real promise of a Content Revolution: not better content management for its own sake, but simpler services, clearer communication and an accessible and inclusive approach to information. And that will create stronger trust and help our government to build better citizen experiences for the long-term.

Success will land with those departments and agencies that know where technology adds value, where people matter most, and how to bring both together for better outcomes for citizens.