Half of the municipal employees are secretly using artificial intelligence
A report by Uniserver, a digital infrastructure company, stated that more than half of municipal employees in the Netherlands (55%) use artificial intelligence tools in their work, without clear policies or official support for using this technology. The report, based on a survey conducted by Motivaction among decision-makers in the field of information technology within municipalities, revealed that 90% of municipalities are concerned about the potential for data leaks or weakened digital security as a result.
Uniserver stated that these figures show "a clear tension: usage is present, but policies are lagging behind." Although most employees use artificial intelligence through personal initiatives, only 55% of municipalities have a clear policy in this area or provide training for employees. Municipalities suffer from a lack of transparency from AI tool developers, as 63% of IT officials indicated that they do not know how these tools handle citizen data and regulatory data. Additionally, 81% expressed doubts about the extent to which the use of artificial intelligence within their municipalities complies with laws and regulations.
Ralph Koutker from Uniserver said: "Artificial intelligence is being used faster than policy development. In municipalities, AI applications often emerge from the ground up: quickly and creatively, but without oversight." He added that this "provides opportunities, but it also carries risks, especially when citizens' data is transferred to systems outside the organization's control."
Many municipalities are trying to improve their conditions and ensure that the use of artificial intelligence complies with regulations, as 64% of them said they plan to update their policies over the next year to include guidelines on artificial intelligence.
Uniserver emphasized the necessity for citizens to have confidence that their municipalities are using their data fairly, transparently, and securely. The company advised against using third-party artificial intelligence tools, preferring to rely on proprietary AI solutions operating within a secure Dutch infrastructure. Kootker stated: "Loss of trust happens faster than building it. Therefore, it is very important to know precisely how citizens' data is used when operating AI."