A wave of states has passed laws requiring disclosure when political communications use AI-generated or manipulated media, and the trend is accelerating. What started as a response to political deepfakes has grown into a patchwork of state rules that reach campaign advertising and, increasingly, direct outreach like texting. Campaigns using AI anywhere in their messaging need to track it.
What do state AI-disclosure laws require?
Broadly, they require that political content which is AI-generated or materially manipulated be labeled as such, especially close to an election. The specifics vary by state, some target deceptive deepfakes of candidates, others require a disclosure on any synthetic political media, and penalties and timing windows differ. The common thread is transparency: if AI made or altered it, the voter should be told.
How does this reach texting?
Directly and indirectly. A text that includes an AI-generated image or video of a candidate could fall under a state's synthetic-media rule, and a message designed to deceive about who's speaking runs into the same concerns that drove the FCC's AI-robocall action. Even AI-drafted text copy, while generally not the target, sits in a tightening environment. The safe read is that any AI-generated media in a political text is subject to disclosure rules in a growing number of states.
What should campaigns do?
Disclose AI-generated media, and keep the sender honest. If you use AI to create images, audio, or video in political messages, label it where any state you touch requires it, and don't use AI to impersonate or mislead about who's speaking. AI used to draft or personalize plain-text messages faster is a different matter, and generally fine, but the moment synthetic media enters a political message, disclosure obligations attach. This fits the broader principle of responsible AI use: work faster, stay honest, identify yourself.
Frequently asked questions
Do you have to disclose AI in political messages?
In a growing number of states, yes, when the content is AI-generated or materially manipulated political media, especially near an election. Requirements and penalties vary by state.
Do AI-drafted text messages need a disclosure?
AI-drafted plain text is generally not the target of these laws, which focus on synthetic media like deepfake images and audio. But the environment is tightening, and deceptive AI content of any kind draws scrutiny.
How should campaigns handle AI disclosure?
Label AI-generated media where any state you text into requires it, and never use AI to impersonate or mislead about who's speaking. Using AI to draft or personalize faster is fine; synthetic media triggers disclosure.
Keep reading: AI and deepfakes in political texting and the FCC AI-robocall ban. For the agency, see the FCC.
This is general information, not legal advice. AI-disclosure laws vary by state and change quickly, current as of the date above; consult counsel.