The new era of public sector service is about building trust through agentic AI

We live in a world where nearly every service is fast, seamless and intuitive. People can verify their identity through a banking app in seconds or call a rideshare instantly — and they expect their government to deliver the same level of speed and efficiency. Waiting days or weeks for a basic government service no longer feels reasonable. Every slow touchpoint, redundant form and broken workflow signals to citizens that the government isn’t keeping pace with the level of service citizens expect. That gap doesn’t just frustrate people; it erodes trust.

This trust gap is not theoretical. The findings from the latest Edelman Trust Barometer show a significant and growing gap in public trust, a clear indication that people’s expectations are rising faster than governments’ ability to deliver.

Public institutions are being asked to operate at private-sector speed with legacy systems, fragmented data and static resources. The workload has exploded; the tools have not. That imbalance has created a structural problem: governments simply cannot meet citizen demand and expectations using traditional models.

The sheer volume of demand is staggering: internal public administration studies in the UK indicate that of the 1 billion annual citizen transactions, about 143 million are "complex repetitive transactions," with roughly 84% being "highly automatable" through artificial intelligence.

We are already seeing how technology can manage this scale. Singapore has already successfully deployed virtual assistants for the public sector. The Government's 'Ask Jamie' has fielded over 15 million queries across 80 government websites, resolving half of the issues that used to clog up call centres.

Similarly, the City of Barcelona has adopted a centralized platform, powered by Salesforce, that gives civil servants a complete, 360-degree view of their interactions with citizens. It allows them to track service delivery and share information across departments to provide a personalized service for each citizen. By connecting previously siloed data, the city has turned reactive bureaucracy into a proactive service, empowering communities with faster, more personalized support. The UK is also harnessing Salesforce’s AI with ‘Bobbi,’ an autonomous AI agent helping three police forces deliver around-the-clock answers to frequently asked questions and assisting citizens in logging cases themselves. In its first week of deployment, Bobbi resolved 82% of inbound queries without requiring escalation to a human officer.

Governments and public administrations have an opportunity to close this trust gap by using agentic AI. This advanced technology powers autonomous intelligence systems or AI agents. These can complete tasks independently, reason through multi-step problems and adapt their actions in real-time to achieve specific goals.

Authorities don't simply give the agent a prompt; they give it a goal, such as 'process this permit renewal.' The AI-powered agent has the autonomy to plan the steps, access the necessary databases, validate the information against regulations and execute the transaction.

Estonia, for example, is piloting 'Bürokratt,' a network of interoperable agents that can cross agency lines. A citizen can ask to renew a passport and the agent coordinates directly with border control to execute the job. The government acts as one cohesive unit, rather than a collection of fragmented departments.

And citizens are clearly ready for this shift. A recent Salesforce survey found that 90% of global respondents are willing to use an AI agent to interact with the public sector. Brazil, Spain, Singapore and Italy are among the most eager to engage, but the appetite is global. In the US and Germany, 20% of respondents say they are already 'very likely' to use an AI agent to help with complex tasks, like filing taxes. The public is ready for a new way of engaging with their governments.

The shift to AI agents could also offer a path to greater transparency, moving administrations from 'black box' governance to 'glass box' governance. But this transparency relies on a critical principle: the human must remain the architect.

By generating a perfect audit log of every rule followed and every action taken, agentic AI empowers the public servant with absolute visibility into the service delivery process. This gives civil servants the tools to review and validate outcomes with clarity, reinforcing their role as the ethical and trusted stewards of the system. The 'glass box' ensures that while the AI executes the transactional work, the human retains full command over the ethics, the guardrails and the final results.

By deploying agents capable of autonomously navigating complex workflows, governments can finally move from just answering citizens' questions to actually solving their problems.

The vision for the next era of government is not about replacing civil servants; it is about empowering them through a hybrid workforce.

We have hit the limit of what human-only bureaucracies can process alone. At Salesforce, we believe the future of public service lies in a 'Humans + AI' model. When agentic AI handles the robotic 'triage' work — sorting forms, checking compliance and routing tickets — it doesn't just clear the backlog; it clears the way for human connection. Public servants are freed to handle the complex, emotional and high-stakes cases that require genuine empathy and judgment, qualities that no algorithm can replicate.

The opportunity now is to scale what already works in the private sector into government. Trust, responsibility and efficiency are the goal, without the need to reinvent the wheel. By combining the speed and scale of AI agents with the experience and empathy of humans, we can build a state that is well-intentioned and highly capable and trusted by its citizens.