As artificial intelligence (AI) becomes embedded into daily life, people are changing how they find information.
Instead of navigating government websites, calling contact centers, or searching through pages of FAQs, residents are increasingly turning to AI assistants and conversational tools to get answers instantly. That shift is raising expectations for every government interaction. According to The State of AI in Government report, 62% of surveyed public-sector employees report their agency is under moderate to extremely high pressure to provide faster, simpler services.
At the same time, the internet itself is changing. New research from Pew Research Center found that approximately 10% of English-language webpages now show significant signs of AI authorship, and more than one-third of webpages published since the release of ChatGPT contain signs of AI-generated or AI-edited content. Further, the prevalence of AI-authored content is continuing to grow.
While residents increasingly expect conversational experiences and immediate answers, governments face an important question: How can agencies ensure citizens receive accurate, authoritative information when more of the internet is being generated, summarized, and reshaped by AI?
The risk of losing control of the answer
Unlike consumer recommendations or general internet content, public service information often has real-world consequences. A misunderstanding about permit requirements, public benefits eligibility, waste collection schedules, business licensing rules, or emergency procedures can create frustration, delay services, and erode public trust.
Historically, agencies controlled this information through official websites, published documents, staff knowledge, and contact centers.
But as residents increasingly ask AI systems for answers, there is a growing risk that the information people receive may not come directly from the agency responsible for providing it.
Instead, those answers may be assembled from search results, outdated webpages, third-party sources, or information that lacks the context and nuance required for government services. How can residents be sure the answers they receive can be trusted?
For governments, trust has always been built on accountability, transparency, and accuracy.
Residents expect information to come from the official source. They expect policies to be interpreted correctly. They expect answers to reflect the latest requirements, regulations, and procedures.
As AI becomes a primary access point for public information, those same expectations remain. In fact, they become even more important.
The future of government AI should not be a black box generating answers from unknown sources. It should be an extension of the agency itself, grounded in official content, aligned with agency policies, and continuously updated as information changes.
That is why governments need a fundamentally different approach to AI than the commercial internet.
From searching for information to accessing trusted knowledge
At Granicus, we believe AI should strengthen trust, not weaken it. That belief is at the core of Government Experience Agent (GXA).
Rather than relying on the open internet, GXA is grounded in agency-approved information. Answers are generated using the content, documents, policies, knowledge bases, and systems that governments already trust and maintain.
When a resident asks a question, GXA does not search the web for the most likely answer. It delivers responses based on the agency’s own authoritative sources.
But trusted AI starts with trusted content. As part of implementation, Granicus helps agencies assess and improve their website’s content readiness by identifying outdated information, closing content gaps, and organizing knowledge so it is easier for both AI systems and residents to find and understand.
This not only improves the quality of responses delivered through GXA but also helps agencies prepare for a future where citizens increasingly discover information through search engines and AI assistants outside of their control.
This approach helps agencies:
- Provide consistent and accurate answers across channels.
- Reduce the risk of misinformation.
- Ensure information stays aligned with current policies and procedures.
- Improve the discoverability of agency information.
- Extend staff expertise through AI-powered self-service.
The result is a better experience for both residents and public servants.
Government AI must be held to a higher standard
The rapid growth of AI-generated content online is a reminder that governments cannot treat accuracy as an afterthought.
When residents have questions about permits, benefits, licensing, public meetings, or community services, they need answers they can trust. Those answers should reflect current policies, official guidance, and agency expertise, not whatever information an AI model happens to find on the internet.
That is why government AI must be transparent, accountable, and grounded in authoritative sources.
As more of the internet becomes AI-generated, agencies have an opportunity to provide a direct connection to trusted information.
By combining AI with authoritative agency knowledge, governments can deliver faster service, improve the resident experience, and maintain their role as the source of truth for the communities they serve.
Learn more about Granicus GXA here.
