Arizona officials use AI version of themselves to warn voters not to be fooled by AI

Arizona Secretary of State Adrian Fontes posted a video ahead of the 2024 election warning voters not to be fooled by AI-generated deepfakes and other fabricated political content ahead of this year's election.

Although it appeared to be Fontes delivering the video message, in fact it wasn't – it was a deepfake of Fontes to drive home the warning and underscore the point that AI-generated content can be convincing even to the untrained eye.

He shared the video during Sunday's episode. Meet the press.

In the video, Fontes' deepfake introduces itself, explains what it is and what its purpose is, before reminding viewers that it's just an “impersonation” of a real person.

“Hello, this is a public service announcement featuring an AI version of Arizona Secretary of State Adrian Fontes. This video was produced as part of the 2024 state election security tabletop exercise,” the AI ​​version of Fontes said. “This was produced with the consent and cooperation of the real Secretary Fontes, but again, that is not me, I am the AI ​​version of him.”

While the video features some rather awkward and robotic dialogue and some puzzling deepfake facial animations, it otherwise stands up relatively convincingly as actual footage of Arizona state officials.

Meet the Press Host Kristen Welker called the footage “horrifying” and asked Fontes why he felt the need to create a deepfake of himself.

“In recruit training, we looked at enemy weapons and trained to counter them as best we could,” he said. “AI is not a new weapon; it's an amplifier and spreader of misinformation and disinformation.”

He said he wanted to ensure election officials had a “good understanding” of how AI could be used to create misinformation and disinformation, and put processes in place to tackle it.

Fontes said the state government had already conducted “tabletop” exercises with media outlets to teach journalists how to spot AI-generated content, and would soon conduct similar exercises with law enforcement officials.

He wasn't the only US official to speak publicly about AI this week. Joe Biden shared his thoughts on the technology on Tuesday, saying AI companies still need to “earn our trust.”

“I am committed to doing everything in our power to advance and demand innovation that is safe, secure, trustworthy and responsible, and that includes the use of AI-generated voice. I ask AI companies to join me in this commitment,” Biden said on X.

Companies developing AI-powered systems are already trying to push them on consumers: Instagram has all but replaced its old search bar with AI search, and Google has started offering “AI summaries” on some searches, whether users ask for them or not.

Google's AI search tool has come under fire after some users reported seeing “hallucinations” in their search results — a “product fail” that is fitting for the AI ​​company's brand.

Gizmodo We've rounded up some of Google's AI-inflicted hallucinations, including one that assures users they can find a recipe for spaghetti with gasoline, one that claims Barack Obama is Muslim, and one that recalls a Spongebob episode in which Sandy Cheeks “commits suicide by overdosing on drugs including cocaine, heroin and alcohol,” which we're pretty sure Nickelodeon would never approve.

Sandy's tragic outcome aside, the search results about Obama's faith are an example of the misinformation generated by AI models.

Disinformation is nothing new in American politics, especially in the age of social media, but given the ease of use of AI products and their ability to create compelling video footage of live politicians, citizens will need to be even more vigilant when searching for facts online.

Or you can embrace the chaos and watch the 24/7 AI-generated Twitch debates between AI Joe Biden and AI Donald Trump. Both are valid.