Chatbots are generating false and misleading information that could disenfranchise voters during the presidential primaries in the US, according to a report by AI experts and election officials. Chatbots are being used by millions of people for basic information about the voting process, but they are prone to suggesting voters head to polling places that don’t exist or inventing illogical responses based on outdated information.

Five large language models tested during a workshop were rated inaccurate and harmful, perpetuating dated and inaccurate information that could limit voting rights. This highlights concerns about generative AI amplifying threats to democracy.