Dutch regulators flagged AI chatbots for unreliability, biasness in voting advice

The Dutch Data Protection Authority (AP), the country’s main watchdog, has issued a strong warning against using AI chatbots when offering voting advice, declaring them to be “unreliable and clearly biased."

The warning has been cited as a particular threat to democracy eight days before the national elections.

The authority said, in a report ahead of the October 29 elections, that the four chatbots tested by the AP “often end up with the same two parties, regardless of the user’s question or command.”

In half of the cases, it was identified that the chatbot proposed either the far-right Freedom Party (PVV) of Geert Wilders or the left-wing GroenLinks-PvdA led by the former European Commission vice-president Frans Timmermans.

The report further states that some parties, such as the Centre-right CDA, “are almost never mentioned even when the user's input exactly matches the positions of one of these parties.”

The deputy head of the AP, Monique Verdier, said that the chatbots may seem like sharp tools but act as a voting aid, they consistently fail.

The far-right party of Wilders is ahead in the polls, but the gap to the GroenLinks-PvDA and CDA seems to be narrowing with many voters yet to make up their minds.

However, all major parties ruled out an alliance with the PVV, which means that the party that comes second is most likely to provide the next prime minister.

What is the next step for chatbots to provide accurate information?

The prime goal is to train and refine these tools to attain reliable results. This requires a multi-faceted approach focusing on strict regulation to enhance transparency.

The role of developers is crucial to reduce societal and political biases to achieve verifiable voter information.

The AP emphasized that the bots were not deliberately biased and that their “identified shortcomings are a consequence of the way AI chatbots operate”.