(Public News Service) – The rise of artificial intelligence is raising alarm bells for election officials in Idaho and across the country. Before the New Hampshire primary in January, a robocall imitating President Joe Biden called voters and told them not to vote. It’s seen as a potential preview of what voters could be in for as the 2024 general election approaches.
Rachel Orey, senior associate director of the Bipartisan Policy Center’s elections project, said while incidences like the one on New Hampshire might be isolated, AI could have other consequences.
“Our bigger concern is what’s known as a ‘liar’s dividend,’ that even when there are instances of generative AI being used to target voters with false information, they feed into this bigger risk that the presence of false information makes voters trust any information less,” she explained.
Orey added the past few years have seen a near constant assault on accurate voting information, which has made it challenging for good information to reach voters. Idaho lawmakers introduced a bill this session that prohibits AI’s use to misrepresent candidates in communications.
Orey said AI could supercharge the misinformation campaigns that have existed for years. However, she added, election officials have a leg up going into the 2024 vote.
“Election officials and voting advocates around the country are sort of well prepared to mitigate and respond to increases in misinformation because they spent the last couple of years flexing that muscle and learning how to respond to misinformation and election denial campaigns,” she continued.
Orey said election officials should have a plan ready to respond to AI misinformation campaigns, which might include contacting affected voters, and added there likely aren’t any regulatory options available at the moment to stop these operations.
“Technology is maybe growing faster than the regulatory tools we have available. So, at present it seems difficult to find the policy that the government could adopt and make these robocalls impossible,” Orey said.
Orey added another concern is targeted campaigns that use a voter’s personal information to persuade them not to vote, although there aren’t any documented instances of this happening yet.