New legislation in Michigan would require disclaimer on deep fakes in political ads
(CBS DETROIT) - Lawmakers in Michigan are concerned deep fakes and the use of AI could have a major impact on our elections.
One Michigan lawmaker wants to make sure you can tell when footage and audio are faked in political advertising.
"Our package of bills requires that if you are going to be using AI to generate political ads or other political material, that you use a disclaimer simply stating that it was generated either substantially or in whole by AI," said State Rep. Penelope Tsernoglou.
It's a technology that has the potential to make a big impact, according to David Derigiotis, a cyber security expert based in Metro Detroit.
"Just think of a couple of use cases on the political front. Seeing the president or leading presidential candidate either affirm their support for a certain nation or to deny support for a certain nation," he said. "Or imagine if you saw the president come on TV and say that he's going to institute a draft. Those are some of the things that we have to be concerned about."
Derigiotis says this technology is easily accessible. He began merging footage of CBS News Detroit anchor Terrance Friday with his own. When this software is done, he can have Terrance say anything he wants for free.
"The technology is really advanced right now so that it's almost impossible to distinguish real images from fake images," said Tsernoglou.
Tsernoglou created her deep fake audio to show her fellow lawmakers.
"Hi, Representative Tsernoglou. It's your buddy, Joe," said a faked clip of President Joe Biden. "I really like your bill that requires disclaimers on political ads that use artificial intelligence. No more malarkey."
"Of course, I've never talked to Joe Biden before," Tsernoglou said.
If passed in Michigan, the legislation would require a written disclaimer on any images or footage created for political advertising. For something like radio, the disclaimer would need to play at the beginning and end of the clip.
"There have to be guardrails around the use of this technology," said Derigiotis. "Like any emerging technology, there can be a lot of great things that happen, a lot of innovation. But there also are certainly abuses that take place with it."