Disclosure alone may undermine trust

A label saying artificial intelligence was used in a political ad can make voters trust the candidate less, according to new research released in April. (editorandpublisher.com) Editor & Publisher reported on April 13 that the American Association of Political Consultants Foundation found a measurable “disclaimer effect” when AI notices appeared on campaign ads. Viewers reported more mistrust and skepticism even when an ad did not actually include AI-generated content. (editorandpublisher.com) The foundation’s summary says the project tested the disclaimer language used in current state laws through a three-phase research design built around realistic campaign ad formats. A related 2024 New York University study with more than 1,000 participants also found that candidates looked less trustworthy and less appealing when ads carried AI disclaimers. (aapcfoundation.org; marketplace.org) Those findings are landing as disclosure rules spread. The Federal Communications Commission proposed rules in July 2024 that would require on-air and written disclosure of AI-generated content in radio and television political ads, and Google said its updated policy would require election advertisers to disclose altered or synthetic content starting in July 2024. (fcc.gov; support.google.com) Outside politics, a local Jacksonville radio report on April 13 showed how easily a synthetic image can still travel. WOKV said an AI-generated picture that appeared to show Jacksonville Jaguars quarterback Trevor Lawrence with short hair fooled some viewers before a team digital content specialist said his hair was still long. (wokv.com) WOKV quoted host Toni Foxx saying she was “duped by AI,” and pointed to a post from Jaguars digital content specialist Marcel Robinson on April 12 rejecting the haircut rumor. The station said the fake image spread because Lawrence’s long hair is a recognizable part of his public image. (wokv.com; blackandteal.com) Some election-law advocates still back disclosure mandates. In a September 19, 2024 filing, Campaign Legal Center told the Federal Communications Commission it supported an AI disclosure rule and urged broad public disclosure across as many media providers and operators as possible. (campaignlegal.org) The split is now clear in public policy and public behavior at the same time: regulators and platforms are adding labels, while new studies and a viral sports fake show that labels do not guarantee people will feel more certain about what they are seeing. (fcc.gov; editorandpublisher.com; wokv.com)