Dutch authorities have warned voters against using AI when voting as it never recommends centrists

The Dutch Data Protection Authority (AP) said AI chatbots are "unreliable and clearly biased" when providing voting advice, warning of such a threat to democracy eight days before national elections, The Guardian reports . The four chatbots the agency tested "often suggest the same two parties, regardless of the user's question or request," according to the report, released ahead of the Oct. 29 election.

In more than half of the cases, the chatbots suggested either the far-right Freedom Party (PVV) of Geert Wilders, or the left-wing GroenLinks-PvdA, led by former Vice-President of the European Commission Frans Timmermans. Some parties, such as the centrist CDA, "are almost never mentioned, even when a user's query closely matches the positions of one of those parties," the report notes.

AP deputy chief Monique Verdier said that while chatbots may seem like smart tools, “as voting assistants they are a consistently disastrous choice.” She added that voters are being pushed towards a party that does not necessarily align with their political views. According to Verdier, this directly affects the cornerstone of democracy - the integrity of free and fair elections, so the department warns against the use of AI chatbots for voting advice, as their work is opaque and difficult to verify.

The Dutch will go to the polls on October 29th in an election that will be closely watched across Europe due to the PVV results. Wilders' far-right party is leading in the polls, but the gap with GroenLinks-PvdA and the CDA has been narrowing recently, with many voters still undecided. All major parties have ruled out an alliance with the PVV, meaning the second-place party is likely to present the next prime minister. AP emphasized that chatbots are not intentionally biased, and “the identified shortcomings are a consequence of how the AI ​​works.”

Bloomberg previously wrote that Meta's ban on political advertising on its social networks in all EU countries in response to the bloc's new transparency rules gave an advantage to right-wing populists. The measure deprived moderate candidates of one of their main ways to reach voters online, while leaving the field open to radical and inflammatory content that platforms' algorithms often promote because of their virality.