Protect Elections from Deceptive AI Act
- Sponsor
- Rep. Johnson, Julie [D-TX-32]
- Committees
- Committee on House Administration (primary)
- Last action
- Sep 10, 2025
Bottom line
This bill aims to combat the use of deepfakes and other AI-generated deceptive content in federal elections by creating a legal prohibition and a private right of action for candidates.
What it actually does
This bill prohibits the knowing distribution of materially deceptive AI-generated audio or visual media featuring federal candidates, if done with the intent to influence an election or solicit funds. It establishes clear definitions for such media, provides exemptions for bona fide news and satire, and creates a civil cause of action for affected candidates to seek injunctive relief or damages.
Proponents argue
Proponents argue the bill is crucial for protecting the integrity of federal elections from the rapidly evolving threat of AI-generated disinformation, which could mislead voters and undermine public trust in candidates. They emphasize the need for clear legal recourse against those who intentionally create and disseminate such deceptive content.
Opponents contend
Opponents might raise concerns about potential infringements on free speech, particularly regarding satire or political commentary, even with the explicit exemption. They might also argue that the definitions of 'deceptive AI-generated audio or visual media' or 'materially deceptive' could be overly broad or vague, leading to chilling effects on legitimate political expression or vexatious litigation.
The bill is concise and can be read quickly, allowing ample time for a thorough understanding of its provisions.
Section 2, adding new Section 325(b) to the Federal Election Campaign Act of 1971.
Prohibition on Distribution of Materially Deceptive AI-Generated Audio or Visual Media
This provision makes it illegal for any person, political committee, or other entity to knowingly distribute AI-generated audio or visual content that is materially deceptive regarding a federal candidate. The distribution must be done with the intent to influence an election or solicit funds.
Supporters argue
Supporters argue this is the core protection needed to prevent the spread of deepfakes and other AI-generated disinformation that could sway elections by creating false narratives about candidates.
Critics contend
Critics might argue that the 'knowingly' and 'intent to influence' requirements could be difficult to prove, making the prohibition less effective in practice. Concerns about free speech implications might also be raised, even with exemptions.
Tradeoffs
Balances the need to protect election integrity from new technologies against potential concerns about free speech and the practical challenges of enforcement.
Section 2, adding new Section 325(a)(2) to the Federal Election Campaign Act of 1971.
Definition of 'Deceptive AI-Generated Audio or Visual Media'
This section defines what constitutes 'deceptive AI-generated audio or visual media.' It includes content created by AI that appears authentic but either fundamentally alters a reasonable person's understanding of a person's appearance, speech, or conduct, or falsely makes a person appear to have exhibited speech or conduct they did not.
Section 2, adding new Section 325(c) to the Federal Election Campaign Act of 1971.
Exemptions for Bona Fide News, Periodicals, and Satire
This provision carves out specific exemptions from the prohibition. It allows radio, television, and streaming services to broadcast such media as part of bona fide news if they clearly acknowledge questions about authenticity. Similarly, newspapers, magazines, and internet publications are exempt if they clearly state the media does not accurately represent the individual. Materially deceptive AI-generated media that constitutes satire or parody is also exempt.
Section 2, adding new Section 325(d) to the Federal Election Campaign Act of 1971, and Section 2(b) of the Act.
Private Right of Action and Defamation Per Se
This provision grants federal candidates whose voice or likeness is used in violation of the prohibition the right to bring a civil action. They can seek injunctive relief to stop the distribution and/or damages. The bill also specifies that a violation of this prohibition constitutes defamation per se, meaning damages can be presumed without needing to prove specific harm.
Requires the plaintiff in any civil action under this section to establish the violation through 'clear and convincing evidence.'
Section 2, adding new Section 325(d)(3) to the Federal Election Campaign Act of 1971.
Why it matters:This is standard legislative practice for specifying evidentiary standards in new causes of action. It's not evasive.
Case for: Proponents would argue that a 'clear and convincing' standard is appropriate for cases involving free speech concerns, ensuring that only genuinely malicious and deceptive content is penalized and preventing frivolous lawsuits.
Case against: Critics might argue that this higher burden of proof (compared to a 'preponderance of the evidence') could make it more difficult for victims to successfully pursue claims, thereby weakening the bill's intended protections against AI-generated disinformation.
Estimated impact: This provision significantly impacts the likelihood of successful litigation for candidates seeking relief under the Act.