New York may have become the latest state to potentially put media outlets that run political ads without AI disclosure in the legal line of fire, but it turns out it is not alone. To date, 11 states have adopted laws that regulate how artificial intelligence can be used in political advertising.
New Mexico’s Campaign Reporting Act went into effect this week. It too says AI-manipulated political ads must carry a disclosure. “This audio has been manipulated or generated by Artificial Intelligence,” says the straightforward label radio ads would need to carry. The New Mexico law says stations will not face legal consequences if they can prove they can certify that they alerted the ad buyer about the AI notification requirements.
In December, Michigan adopted a new law that requires a campaign to disclose when AI technology was used to make an ad. The radio disclaimer would need to state: “This message was generated in whole or substantially by artificial intelligence.” For radio ads, the statement would have to be at least three seconds in length and spoken in a clearly audible and intelligible manner in the same language as the rest of the ad at either the beginning or end of the commercial. Another new Michigan law prohibits the use of AI-generated deepfakes within 90 days of an election.
As is the case in other states, in Michigan stations must play the role of informer. There the law says media outlets are not liable if they disclose AI use — as long as they can show they provided notice to the ad buyer about the requirement to label ads that use it.
There are several other states where there are now limits on how AI audio, video and images can be used in political ads. The other states adopting new laws during the past year include California, Idaho, Indiana, Minnesota, Oregon, Texas Utah, Washington, and Wisconsin. Many others are considering similar legislation.
Broadcast attorney David Oxenford says that while most states have some protections, not all do, pointing out in a blog post that Minnesota is one state where AI disclosures could leave media outlets liable. But he says the biggest concerns are the significant compliance issues raised by the state laws mandating ads include AI disclosures when they air.
Under federal law and FCC rules, he points out that broadcasters are prohibited from censoring any ads from candidates or their authorized committees. It is why FCC and the courts have even required broadcasters to air ads with graphic or racist content. “Thus, the no-censorship provision could require broadcasters to run candidate ads containing AI even if the ads don’t comply with the labeling requirements imposed by a state statute,” Oxenford writes.
What may become a bigger issue in the coming years is how stations will deal with the near-impossible task of figuring out what has been generated by AI. “Many may shy away from all political advertising to avoid being put in that position,” Oxenford predicts, noting that broadcasters have no choice to run ads from federal candidates.
Comments