Fraudulent Artificial Intelligence Regulations (FAIR) Elections Act of 2026
- Sponsor
- Sen. Merkley, Jeff [D-OR]
- Committees
- Rules and Administration Committee (primary)
- Last action
- Jun 11, 2026
Bottom line
The FAIR Elections Act aims to safeguard election integrity by combating AI-driven disinformation and strengthening voter registration protections against erroneous removals, while restricting federal data use in state voter eligibility determinations.
What it actually does
This bill establishes federal prohibitions on distributing false AI-generated election media intended to impede voting or harass election officials, with exemptions for news organizations. It also amends the National Voter Registration Act of 1993 to restrict states from removing voters from official lists based on unverified databases, requiring specific governmental or Attorney General-approved data sources. Furthermore, it prohibits the Federal Government from sharing its data with states for voter eligibility comparisons and mandates a GAO study on the use of federal databases for voter roll maintenance.
Proponents argue
Proponents argue that this bill is crucial for protecting democratic processes from emerging threats like AI-generated disinformation and for preventing eligible voters from being disenfranchised by inaccurate or politically motivated voter roll purges. They emphasize the need for robust federal standards to ensure fair access to the ballot and secure elections in the digital age.
Opponents contend
Opponents might argue that the bill's restrictions on AI-generated media could infringe on free speech, particularly satire or political commentary, and that the federal government's role in dictating state voter list maintenance procedures oversteps states' rights and could hinder efforts to ensure accurate voter rolls by removing ineligible voters. They may also raise concerns about the privacy implications of requiring specific data sets for voter verification.
The bill is relatively short and its core provisions are clearly stated, making it accessible for an informed reader to evaluate within a reasonable timeframe.
Section 2
Prohibiting False AI-Generated Election Media
This section makes it illegal to knowingly distribute AI-generated text, images, audio, or video that contains materially false information about election logistics (time, place, manner of voting, or voter qualifications) with the intent to impede voting, or that falsely depicts an election official with the intent to intimidate them or deter voting. It provides exemptions for news organizations that clearly disclose the falsity.
Supporters argue
Supporters argue this provision is a necessary and timely response to the growing threat of AI-powered disinformation, which can rapidly spread false information and undermine public trust in elections, potentially suppressing voter turnout or endangering election workers.
Critics contend
Critics might contend that this provision could be overly broad, potentially chilling legitimate speech, satire, or political commentary, and that determining "materially false information" or "intent" could be subjective and lead to selective enforcement. Concerns about free speech implications are often raised.
Tradeoffs
This provision balances the need to protect election integrity from malicious AI-generated disinformation against concerns about potential restrictions on free speech and the challenges of defining and enforcing "materially false" content and "intent."
Section 3, inserting new Section 8A into the NVRA.
Limiting State Voter Roll Purges to Verified Data Sources
This section prohibits states from removing voters from official lists or challenging voter eligibility unless the determination is based on specific, verified source information. This includes data from governmental entities obtained in official duties or data approved by the Attorney General that meets strict criteria, such as containing "qualifying data sets" (e.g., full SSN, driver's license number, or last four SSN plus DOB) and being updated monthly.
Section 4
Restricting Federal Government's Use of Data for State Voter Eligibility
This section explicitly prohibits the Federal Government from comparing its internal data on individuals with state or local government data for the purpose of determining voter eligibility in federal elections. It also creates a private right of action, allowing individuals aggrieved by a violation to sue for declaratory or injunctive relief.
Section 5
GAO Study on Federal Database Impact on Voter Rolls
This section mandates the Comptroller General of the United States (GAO) to conduct a study on the effects of using federal databases, such as the SAVE system, for removing individuals from voter rolls. The study must examine whether such uses suppress voting rights, the accuracy of the identification process, and the number of incorrect removals in each state. A report is due to Congress within 180 days.
Exempts bona fide news organizations (radio, TV, cable, satellite, streaming, newspapers, magazines, internet publications) from the prohibition on distributing false AI-generated election media, provided they clearly acknowledge the information is false.
Section 2(c) Inapplicability to Certain Entities.
Why it matters:Standard legislative practice to define exceptions and carve-outs within a prohibition.
Case for: This exemption protects journalistic freedom and the ability of news outlets to report on or analyze false information, even if AI-generated, as long as they are transparent about its nature. It prevents the law from being used to suppress legitimate news coverage.
Case against: Some might argue that even with a disclaimer, broadcasting or publishing false AI-generated content, especially close to an election, could still contribute to confusion or be weaponized, regardless of intent. They might suggest stricter rules for all distributors.
Estimated impact: Significant, as it defines the boundaries of the prohibition, allowing established media to discuss or show false AI content for informational purposes without legal penalty, provided proper disclosure.
Defines "qualifying data sets" for voter verification, requiring specific combinations of personal information like full social security number, driver's license number, or date of birth.
Section 3(b)(2) Qualifying data sets.
Why it matters:Standard legislative practice to define technical terms precisely.
Case for: Proponents argue that requiring specific, robust data sets ensures a high degree of accuracy in voter verification, reducing the likelihood of erroneous removals and protecting eligible voters. It sets a clear, verifiable standard.
Case against: Opponents might argue that requiring such specific and sensitive data (like full SSN) could raise privacy concerns, create barriers for states that don't collect or have access to all these data points, or make it harder to cross-reference with existing state databases that use different identifiers.
Estimated impact: High, as it directly dictates the type and quality of data states must use for voter roll maintenance, potentially requiring significant changes to state systems and data sharing agreements.