AI in Government Customer Service: Don't Cut Staff Yet
Gartner predicts many AI-driven customer service layoffs will be reversed. Here's what state and local government contact center leaders and managers should do instead.
Rosetta Lue
9/25/2026
AI in government customer service: why cutting staff first could cost you more
Private companies are learning a hard lesson about AI. Now it's government's turn to learn it without paying the same price.
Gartner now predicts that 30% of workers laid off in AI-driven cuts will need to be rehired by 2029. Often at a higher cost. And customer service is leading the way.
Emily Potosky, a senior director analyst at Gartner, told CX Dive that contact center staff were "hit first and maybe hardest by the AI hype." In February, Gartner went further. It predicted that half of companies that cut customer service staff because of AI will rehire people by 2027 to do similar work.
If you run a 311 center, a state agency call center, or any public-facing service desk, pay attention. This story is about you.
The private sector already tried this
Klarna is the best-known example. The company said its AI chatbot could do the work of 700 customer service reps. More than a year later, it started hiring human reps again.
The money isn't adding up either. According to Gartner, service and support teams spent a median of almost $6 million on AI in 2025. Of the AI use cases Gartner reviewed, about one in four lost money. About one in four made money. And 42% had no clear value at all.
So most AI projects in customer service can't yet prove they pay off. That should give every city manager and agency head a reason to pause.
Why the risk is bigger for government
A company that cuts too deep can post a job opening next week. Government can't move that fast.
Think about what it takes to rehire in the public sector. Budget cycles. Hiring freezes. Civil service rules. Union agreements. Background checks. Training that takes months. By the time you rebuild the team, residents have been waiting on hold for a year.
And residents can't switch providers. When a bank's chatbot fails, a customer can take their money somewhere else. When a city's chatbot fails, a resident just gets stuck. They still need their pothole fixed, their trash picked up, and their benefits question answered.
That makes AI in government customer service a trust issue, not only a budget issue. Every failed interaction chips away at public confidence.
The better path: reshape the workforce
Gartner's advice is simple. Don't think "replace." Think "reshape."
Potosky said leaders who want real return on AI need to focus on workforce reshaping, not flat headcount cuts. People won't be fully replaced. Their jobs will change.
Forrester principal analyst Max Ball described what that change could look like. Frontline reps could become "bot unblockers" who manage teams of AI agents. Senior reps could act as judges and experts, handling cases too complex for AI.
Companies will also need new roles to keep AI working, including business solutions architects and knowledge managers.
That last role matters a lot for government. AI is only as good as the information behind it. If your knowledge base is outdated, your chatbot will give residents wrong answers, just faster. Someone has to own that content every single day.
Five questions to ask before you cut a single position
Before your agency ties AI to staff reductions, get honest answers to these:
Has the AI proven itself on real resident calls? Pilot results in a demo are not the same as a busy Monday after a storm.
Who will keep the knowledge base current? Program rules, fees, and hours change. Your AI needs a human owner.
What happens when the AI gets stuck? Residents need a clear, quick path to a live person.
Do you know your true AI costs? Potosky pointed out that it's still unclear whether AI prices will go up. Build that into your budget.
How fast could you rehire if you had to? If the answer is "not fast," move carefully.
Measure service, not just savings
The smartest agencies judge AI by how well residents are served. Are first-contact resolution rates going up? Are complaints going down? Are vulnerable residents still able to get help?
Standards like ISO 18295 give contact centers a clear way to measure service quality. Use them. They keep the focus on outcomes instead of headcount.
How GovCXP Digital Partners can help
At GovCXP Digital Partners, we help state and local governments plan for AI in a way that protects both residents and the workforce. We work with 311 and government contact center leaders on AI readiness, knowledge management, workforce planning, and service standards.
The private sector is showing us what happens when AI is treated as a shortcut to cutting staff. Government doesn't have to repeat that mistake.
Ready to build an AI plan that holds up? Contact GovCXP Digital Partners to talk about where your contact center stands today.
Source: Kristen Doerer, "Are AI-driven customer service cuts here to stay? Gartner predicts rehiring," CX Dive, September 11, 2026.