(Source: MIRS.news, Published 11/09/2023) Groups that fail to clarify that artificial intelligence (AI) was used to create a qualified political advertisement could be penalized with a 93-day, $1,000 misdemeanor for their initial offense, under legislation approved by the Senate Thursday and enrolled by the House.
House Elections Committee Chair Penelope Tsernoglou (D-East Lansing) sold the bills to a Senate committee on Wednesday as needing to be passed and in effect by the 2024 election to prevent voters from being fooled by images that are complete computer fabrications that are exceedingly difficult, if not impossible, to spot.
Before passing the bills, however, Sen. Mallory McMorrow (D-Royal Oak) offered exceptions to businesses or business affiliates regulated by the Michigan Public Service Commission (MPSC) or the Federal Communications Commission. They additionally provided guidance for directing viewers to unedited versions of altered pre-existing images, audio or videos.
Another Senate amendment tightened up AI's definition so content typically accepted as political hyperbole isn't unintentionally wrapped into the statute, assuming the Governor signs it.
HB 5141, HB 5143, HB 5144 and HB 5145 were approved Thursday by party-line votes, 20-16.
Under HB 5141 by the House Elections Chair Penelope Tsernoglou (D-East Lansing) specifically, a "qualified political advertisement" would include political advertisements made by or on behalf of a federal, state or local candidate or a committee behind a ballot question.
If their advertisement featured any image, audio or video generated either substantially or completely through AI use, the creator would be responsible for clarifying that. A third or subsequent violation would result in a two-year, $2,000 felony.
Different penalties would be placed on non-qualified political ads, where an election, candidate or ballot question is mentioned. The first offense for lacking a disclaimer in the non-qualified ads would result in a maximum civil fine of up to $250.
