Artificial intelligence (AI) has been utilized in several aspects of the elections with both positive and negative use cases. These use cases continue to expand, among them recently AI chatbots as a means for voters to identify the qualifications and platform of electoral candidates that match their interests, some even directly asking these chatbots who to vote for becoming more integrated into how elections are shaped.
Several studies have looked into the voters’ use of AI chatbots in seeking information about candidates and political parties, if not directly asking who to vote for, as part of the electorate’s decision making process.
Users resonated with the AI chatbots for their straightforward and understandable responses and the interaction that allows them to further probe on the outputs. Users also deemed the responses to be ‘accurate’ and ‘politically neutral’.
Experts, however, have cautioned users against inaccuracy and biases in the responses of AI models. In a news article, Cornell University professor of information science, marketing and psychology David G. Rand said that AI tends to reaffirm and mirror users’ biases, framing candidates’ views through the voters’ lens, rather than objective facts.
In this issue of “Elections and Technology Reader”, ANFREL looks into the use of AI chatbots as a source of information and platform for candidate evaluation to aid electoral decision making for voters and its implication to electoral integrity. This issue also presents studies on user evaluation of AI chatbots, accuracy and potential bias of AI models as well as AI usage policies as well as policies in place on use of AI and other emerging technologies during elections.
Read the full issue here: “Would you ask AI who to vote for?”
