Dutch voters say they'll skip AI at the ballot box — but are they already using it without realizing?

This explores how many Dutch voters say they intend to ask AI for voting advice, and whether that figure tells the whole story. The headline number in the collection comes from an AlgoSoc survey of 1,220 Dutch adults. Scaled up to the national population, it suggests about 1.8 million people might consult AI on how to vote How many Dutch voters might seek AI voting advice?. Willingness drops sharply with age, and younger voters show the most interest. The researchers also flag that actual use may be higher than stated intent: people often use a chatbot casually without thinking of it as 'getting voting advice.' So 1.8 million is better read as a floor than a ceiling.

An American comparison puts this in context. A Change Research survey of 1,892 registered US voters found only 15% likely to use AI chatbots for election information, against 68% for news articles Will voters actually use AI chatbots for election information?. The pattern is the same: use is concentrated among younger voters, and many say they distrust AI's accuracy. Globally, the Reuters Institute finds that about 10% of people use AI chatbots for news. Trust predicts that use more strongly than it does for social media: 44% of chatbot users trust the news they get from them, against 17% of non-users Does trust in AI chatbots drive news-seeking behavior?. Put together, AI election use is a minority habit, but one that grows as people come to trust it.

The part you might not expect is what those 1.8 million would actually be told. The Dutch Data Protection Authority tested general-purpose chatbots as voting guides. In more than half of cases they recommended one of just two parties, PVV or GroenLinks-PvdA, even when the user's stated positions matched a different party Do chatbots steer Dutch voters toward the same parties?. Centrist parties came up less than 2% of the time. The authority called this a 'vacuum cleaner effect' that collapses a multi-party system into two poles Do AI chatbots systematically bias voters toward extreme parties?. The cause traces back to chatbots drawing on unstructured internet text rather than on structured party-position data like the Netherlands' established voting-advice tools use. In a country with many parties, the problem is less about wrong facts than about which parties get skewed or left out.

Getting facts right doesn't fix this. In the US, States United found ChatGPT and Google's AI reached zero verifiable factual errors on voter questions by early 2026. Even so, they pointed people to official election sites less than half the time Do AI chatbots give voters accurate election information?. A chatbot can be accurate about every fact and still steer people toward a narrow set of choices.

So the honest answer is this. Roughly 1.8 million Dutch people say they might use AI for voting advice, probably more do in practice, and the corpus has no harder figure than that one survey. The more useful question may be what kind of advice they get. That depends on whether the tool is built on structured political data, and on whether people treat its answer as one input among several rather than a verdict Should AI outputs replace or supplement human judgment?.