AI assistants have been shown to be “inaccurate and evidently prejudiced” when providing election recommendations, the Netherlands’ data watchdog has reported, issuing warnings about a possible risk to democracy just over a week before the country’s general election.
Study Shows Consistent Bias in Chatbot Recommendations
The four chatbots evaluated by the AP “often end up with the same two parties”, irrespective of the user’s question or command, the regulator explained in a analysis published ahead of the 29 October election.
In more than half of the cases, the chatbot suggested either the right-wing PVV of Geert Wilders or the leftwing GroenLinks-PvdA led by the former European Commission vice-president Frans Timmermans.
Certain groups, such as the centre-right CDA, “are rarely ever cited”, despite the fact that the voter’s query precisely aligns with the positions of these parties, the report pointed out.
Authority Figure Cautions Against Relying on AI Assistants for Voting Guidance
The vice-chair of the AP, Monique Verdier, stated that while chatbots may seem like clever tools, “as a voting aid, they consistently fail”. Citizens were being steered toward a party that might not correspond to their ideological beliefs, she added.
“This directly impacts a fundamental pillar of democracy: the honesty of open and just votes,” said Verdier. “We therefore warn against using AI chatbots for voting advice, as their operation is unclear and hard to confirm.”
Election Context and Party Dynamics
The Dutch head to the polls on October 29 in an election that will be closely watched around Europe for the showing of the PVV.
The right-leaning group of Wilders is ahead in surveys but the lead over the GroenLinks-PvdA and CDA seems to be closing, with many voters yet to make up their minds.
All major parties have ruled out an alliance with the PVV, meaning the runner-up party is highly probable to supply the future head of government.
Inadvertent Prejudice Attributed to AI Operation
The AP emphasised that the bots were not deliberately biased, and that their “noted flaws” are a result of the manner in which AI assistants function.
