PHOENIX (CN) — Arizona state leaders say they’re standing guard against potential election threats fueled by AI, but just weeks away from the general election, they haven’t seen much.
In a two-hour panel on Thursday evening, Arizona Secretary of State Adrian Fontes and other local leaders discussed the potential impacts of AI-generated deep fakes and misinformation on election integrity, and opined on potential solutions.
The panel was co-hosted by Arizona State University’s McCain Institute and political advocacy group Alliance for Securing Democracy.
“We haven’t seen as much AI in 2024 as a lot of us expected,” Bret Schafer, a senior fellow at the Alliance for Securing Democracy, said to the audience in an ASU lecture hall. “That leaves me a little bit optimistic, though I don’t think that means we should let our guard down.”
Fontes said election workers across the state have been training to recognize deep fakes and other AI misinformation campaigns so they can help voters do the same. He said election workers are run through as many different circumstances as possible to be ready for a potential threat.
“We want to be prepared for just about any scenario,” he said. “It’s kind of like Dungeons and Dragons for election nerds.”
In December, his office created AI deep fake videos of himself spreading misinformation to see if election workers would catch it.
Not only are election workers trained to help voters navigate misinformation, he said, but they are trained to recognize AI-generated attempts to scam or hack election sites that might expose sensitive information or jeopardize election security.
“We’re intercepting phishing attempts all the time,” Fontes said.
So far, Fontes said he hasn’t seen much as far as disinformation campaigns, but he warned voters to stay vigilant, and to remain skeptical of everything they read.
“We need to be a little bit more responsible as consumers of information,” he said. “If something feels a little weird, chances are it is.”
Panel moderators wondered whether public policy should be put in place to regulate AI use, but Maricopa County Supervisor Bill Gates said he’s hesitant to avow anything regulating free speech.
“I’m sure there will be policy prescriptions, but I don’t think we know yet what policy and laws ought to be passed,” he said.
Arizona has tried before to regulate AI, but was unsuccessful. The Republican-majority Senate introduced two bills this year criminalizing the creation of fake images, video or audio recordings with the intent to exploit or defraud. Arizona Governor Katie Hobbs vetoed one of the bills and promised to sign the other, but by the time the Legislature concluded its session, the bill remained signatureless on the Democrat’s desk.
When asked about AI-generated malware, Gates assured voters that ballot tabulation machines are not connected to the internet, and therefore cannot be influenced by sources outside official election sites.
Schafer said the threats of misinformation posed by AI are “nothing new.” The difference now is how quickly misinformation can be generated and spread across social media networks before it’s detected as such.
“The adversaries are gaming the system,” Schafer said. “It’s the drowning out of the positive information that concerns me.”
Regardless of potential threats, panelists remained confident in the integrity of Arizona elections.
“They’re run by good people who are well intentioned,” Fontes said. “I believe in them. They have consistently run great elections in Arizona.”
“There’s no proof otherwise,” he added.