LANSING, Mich. (WZMQ) – New legislation currently being reviewed in Michigan’s Senate would create mandated labels for campaign-related content created using artificial intelligence.
The creation of media using artificial intelligence is a new but rapidly growing field. Schools have been quick to create strict rules about the use of AI programs, and just last week, President Biden announced an executive order requiring safety, civil rights, and labor assessments be done to determine the impacts of AI.
Legislation now being reviewed by Michigan’s senate would require any audio, video, or images created using ai to be labeled specifiying its use. East Lansing Representative Penelope Tsernoglou introduced the bi-partisan bills last month because of a rising concern surrounding intentionally deceptive AI-generated media, and deep fakes.
“This type of technology can be used to spread misinformation all the way up and down the ballot.” Rep. Tsernoglou said. “I think voters should keep themselves up to speed with things, and they should do research and their homework, they absolutely should. But they also shouldn’t have misinformation put out there when it looks like it’s real and there’s no way to tell that it’s not real.”
Together, house bills 5141, 5143, and 5144 would restrict the use of artificial intelligence and manipulated media in political campaigns and would generally require campaign advertisements and other political media to disclose of the use of artificial intelligence systems. house bills 5142 and 5145 are companion bills to hbs 5141 and 5144, respectively, that would make complementary changes to the code of criminal procedure… [The bills would} incorporate the proposed new felony of failing to disclose the use of artificial intelligence in political communication. Under the bill, a third or subsequent offense would be a class g felony against the public trust punishable by a statutory maximum term of imprisonment of two years.Legislature.mi.gov – Legislative Analysis
Deep fakes are videos of people generated using AI, or similar technology, that looks like a person and sounds like the person, but it is not. Representative Tsernoglou says the unregulated use of AI could be devastating for political candidates, and races. She said that deep fakes can be used to misrepresent a candidate and their views to distort opinions and potentially change the outcome of elections.
Just this April, the Republican National Committee released an ai-generated video showing made-up future news coverage from a scenario where President Biden is reelected. She says the goal is to create policy that would require similar ads to disclose the use of AI. Representative Tsernoglou explained this bill isn’t an outright ban but would create label requirements similar to those that already exist for sponsored ads.
“The bills require that if you’re doing a political ad that you put. a disclaimer on there that says if it’s substantially or wholly created by AI. ” Tsernoglou said. “Everyone in the campaign world knows you have to put a ‘paid-for by’ disclaimer on your yard sign. This is really comparable to something like that.”
Rep Tsernoglou said she hopes to see this legislation passed soon to ensure people are aware of potentially misleading information ahead of next year’s presidential elections.