Over half of municipal workers are using AI, but policy is lagging behind

Over half of municipal employees (55%) use AI tools at work, with no formal guidelines or support on using the technology. Almost all municipalities (90%) are worried that this will lead to data breaches or inadequate data security, digital infrastructure company Uniserver reported after commissioning Motivaction to survey IT decision-makers in Dutch municipalities.

“The figures reveal a clear tension: AI is being used, but policy is lagging behind,” Uniserver said. While more than half of municipal workers use AI tools on their own initiative, only 55 percent of municipalities report having developed a policy for AI use and supporting their employees with training or education.

At the same time, 90 percent of IT decision-makers working for municipalities are worried about data leaks and inadequate data security. This is partly due to a lack of transparency from AI developers. 63 percent of IT decision-makers say they don’t know how AI tools handle citizen and organizational data, and 81 percent doubt whether AI use in their municipality sufficiently complies with laws and regulations.

"AI use is growing faster than policy. In the municipal sector, we often see AI applications emerge from the bottom up: fast, creative, but uncontrolled," says Ralph Kootker of Uniserver. "This creates opportunities, but also risks. Especially when citizen data ends up in systems outside the organization's control."

Many municipalities are working on improvements and ensuring that their AI use aligns with legislation. 64 percent said they want to update their policy with AI guidelines by next year.

Citizens must be able to trust that municipalities are using their data fairly, transparently, and securely, Uniserver said. The company therefore advised against using AI from third parties, instead opting for private AI solutions running within secure, Dutch infrastructure. “Losing trust happens faster than building it. That's precisely why it's so important that they know exactly what happens to citizen data when AI is deployed,” Kootker said.