Stories about “deepfakes,” “synthetic media,” and other forms of artificial intelligence being used in political campaigns, including in advertising messages, have abounded in recent weeks. There were stories about a superPAC running attack ads against Donald Trump where Trump’s voice was allegedly synthesized to read one of his tweets condemning the Iowa governor for not supporting him in his Presidential campaign. Similar ads have been run attacking other political figures, prompting calls from some for federal regulation of the use of AI-generated content in political ads. The Federal Election Commission last month discussed a Petition for Rulemaking filed by the public interest group Public Citizen asking for a rulemaking on the regulation of these ads. While the FEC staff drafted a “Notification of Availability” to tell the public that the petition was filed and to ask for comments on whether the FEC should start a formal rulemaking on the subject, according to an FEC press release, no action was taken on that Notification. A bill has also been introduced in both the Senate and the House of Representatives to require that there be disclaimers on all political ads using images or video generated by artificial intelligence revealing that they were artificially generated (see press release here).
These federal efforts to require labeling of political ads using AI have yet to result in any such regulation, but a few states have stepped into the void and adopted their own requirements. Washington State recently passed legislation requiring the labeling of AI-generated content in political ads. Some states, including Texas and California, already provide penalties for deepfakes that do not contain a clear public disclosure when used in political ads within a certain period before an election (Texas, within 30 days and California within 60 days).