THE HAGUE: The Netherlands’ Data Protection Authority (AP) has issued a stark warning that consulting AI chatbots for voting advice poses a significant threat to democracy.
The Dutch watchdog has urged the public not to trust artificial intelligence with the critical decision of who to vote for.
This warning follows an experiment that revealed disturbing results. Four different chatbots tested by the authority predominantly favored parties on either the left-wing or the far-right of the political spectrum.
The report clarified that this bias is not believed to be intentional, but is rather an inherent result of how these AI models function. Regardless of the cause, the consequences could be severe. The authority warned that this flawed and biased advice from AI could directly interfere with the fundamental democratic principle of free and fair elections.
