Artificial intelligence; require disclosure when used in political advertisements.
SB 2642 amends Mississippi’s election-disclosure law to require political advertisements that use artificial intelligence to clearly disclose that fact to the public. The bill defines a “qualified political advertisement” broadly to include certain campaign ads for candidates or ballot questions that contain AI-generated image, audio, or video content and are made by or on behalf of a candidate, committee, or other person. It also defines artificial intelligence for purposes of the statute and sets out detailed disclosure rules depending on whether the ad is text/graphic, audio, video, or a prerecorded telephone message.
The bill requires the disclosure to be clear and conspicuous and, in some cases, to specify what portion of the content was produced by AI. For audio and video ads, the disclaimer must be spoken and timed to remain audible or visible for a minimum period. The measure also preserves existing campaign disclaimer rules identifying who paid for or approved the message, and it creates exceptions for certain broadcasters, news coverage, satire or parody, and some distribution platforms that have compliant disclosure policies. The act takes effect July 1, 2025.
SB 2642 would amend Section 23-15-897 of the Mississippi Code by adding AI-specific disclosure requirements to the state’s campaign materials and political advertising rules. It would impose new labeling obligations on candidates, political committees, and other persons who create or distribute qualifying political ads using AI, while also establishing civil penalties for noncompliance and authorizing injunctive relief actions by the Attorney General or an injured candidate. The bill would affect campaign advertisers, political committees, broadcasters, online platforms, and vendors producing political content, and it would create separate violations for each unlawful distribution or airing of an ad.
The available record shows no committee debate, recorded votes, or formal opposition in the provided materials, so the bill’s overall sentiment cannot be measured from discussion history. Based on the caption and text, the measure appears to be framed as a transparency and election-integrity proposal rather than a partisan policy change. Its structure suggests an effort to address public concern about deepfakes and deceptive AI-generated campaign content while preserving ordinary campaign disclaimer practices.
The main potential points of contention are the scope and enforceability of the AI disclosure mandate. The bill reaches a broad range of political communications, including search ads, display ads, video, native ads, sponsorships, and prerecorded calls, which could raise concerns from campaigns, vendors, and media platforms about compliance burdens and the risk of penalties for repeated distribution. Another likely issue is the breadth of the definition of AI and whether the disclosure requirements could sweep in partially AI-assisted content or create ambiguity about what must be disclosed. Exemptions for broadcasters, news organizations, satire/parody, and platforms with disclosure policies may also draw scrutiny over whether they are sufficient or unevenly applied.