Building trust and scaling responsible AI in government

Michael Bostian didn’t set out to build a career that would bring together artificial intelligence, public trust, and government operations. But looking back, a clear throughline has defined his work: applying technology thoughtfully to help people navigate the moments that matter. In his leadership role at Maximus, Michael identifies where digital and AI-enabled solutions can responsibly improve how state and local government programs operate — especially within highly regulated environments where accuracy, accountability, and trust are key.

Early in my career, I worked on large-scale digital transformation efforts in the financial services sector, including the push toward self‑service and automation. What stood out wasn’t just the adoption rate of self-service tools like mobile check deposit — it was the why. People wanted flexibility. They wanted to handle important tasks on their own time, in their own space, without standing in line or rearranging their day. Those expectations have only grown and now extend to how people want to interact with government programs, too.

That transformation shaped how I think about innovation. It’s not about moving fast for the sake of speed. It’s about applying technology to the right use cases and helping people navigate complexity without losing trust along the way.

The government has a fiduciary responsibility to taxpayers. Budgets are fixed, investments need to last, and mistakes are costly to correct. That reality requires more deliberate decision‑making — especially when it comes to AI, and that’s a good thing.

Public agencies can’t afford to chase hype cycles or deploy technology simply because it’s new. A deliberate pace creates space to think critically about where AI truly adds value and how it should be governed over time.

We’re seeing a lot of state programs with AI pilots driven by curiosity, experimentation, and learning, which is healthy. What matters next is being selective about what works and scaling those solutions responsibly. That’s where human-in-the-loop design becomes critical.

The question isn’t whether humans should be involved — they should be. It’s where human judgment adds the most value. Some tasks are routine and can be automated safely. Others are complex and require discretion. Responsible AI means being intentional about how work is divided, so you can build confidence and trust without creating unnecessary friction or slowing progress.

That balance isn’t static. As AI capabilities evolve and confidence grows, agencies should continuously reassess how responsibilities are shared between people and technology.

Demand for government services isn’t decreasing. Populations are growing, interactions with public programs are increasing, and agencies are facing real workforce constraints. In that environment, AI becomes essential for maintaining service levels and meeting public expectations.

The conversation about AI in the public sector needs to shift from replacement to task redesign. The real opportunity is improving throughput, speed, and efficiency in how agencies interact with the public.

When it’s used thoughtfully, AI can be a force multiplier. It can help people more easily find information, navigate complex processes, and complete tasks that otherwise require significant manual effort from staff. At the same time, it allows program teams to dedicate more attention to complex cases where human insight and empathy matter most.

One area where the private sector consistently excels is proactive communications — reaching people at the right time with the right information. I’m not suggesting government should behave like a retailer, but there are moments when residents need timely, precise guidance. 

AI creates an opportunity to provide that guidance responsibly by helping people understand what services are available, what actions they need to take, and what happens next. But technology alone isn’t enough. Agencies also need to invest in strong change management. Alignment around training, communication, and process redesign is just as important as the tools themselves.

Pairing trust with technology

At the core of Michael’s work is a simple principle: technology should respect people’s time, protect their data, and help them move forward with confidence.

“The best government experience,” he says, “is completing a task successfully, and trusting that it’s truly done.”

That belief continues to guide how Michael approaches innovation at Maximus, connecting advanced technology, thoughtful design, and human judgment to build solutions that earn and sustain public trust.