If you were planning to post an edited photo online of a Virginia political candidate during the next election, you might’ve been in trouble.
After FIRE’s opposition and outreach on this bill, Gov. Glenn Youngkin just prevented that from happening by vetoing HB 2479.
The Virginia General Assembly passed HB 2479 to suppress “altered” and AI-generated depictions of candidates — enforced with threats of fines and even jail time — unless a conspicuous disclaimer was added. Instead of trusting the public to decide what’s true, false, or credible, HB 2479 would have violated the free speech rights of Virginians to make the government into the arbiter of truth.
This bill would’ve made it illegal for virtually any Virginian to sponsor an “electioneering communication” that contains “altered” or AI-generated images or audio recordings of identifiable candidates running for elected office. This included messages appearing in print, TV, radio, or online platforms within 60 days of an election.
Not only would it have included traditional paid campaign ads, but anyone’s speech expressing support for or against a candidate that involves the exchange of something of value and appears in a paper, a broadcast, or is promoted online for a fee. This could include using an AI tool that requires a paid license or even posting on a social media platform using a paid premium account that many platforms offer to extend the content’s visibility and reach.
What “altered” means is anyone’s guess — but the government would be the decider. Any edit that created a “fundamentally different impression” of the photo or video could count, meaning it could have covered even simple edits like cropping a photo. If an image of a candidate was cropped to fit onto a page, an aggrieved candidate could sue and argue that the crop created a “fundamentally different impression” from the original if the portion cropped out removed some kind of context — such as part of the background or another person.
And every speaker was covered, not just mud-slinging political opponents. Suppose a small business owner buys space in a newspaper to highlight how a mayor running for reelection failed to address public safety concerns outside her shop. If she includes a slightly edited and unflattering image of the mayor, she could have been sued — even if the content is not misleading (or even relevant).
The disclaimer requirement wouldn’t have solved the bill’s problems, and in fact created new ones. The First Amendment protects both your right to speak your mind and to hold your tongue, but disclaimers force you to utter government-mandated speech. Even worse, the disclaimer here could have actually misled voters into thinking that someone is spreading falsehoods — even if the ad was factually accurate — simply because edited or AI-generated material was included.
Lawmakers certainly need to protect the electoral process, but this bill would have done the opposite, and it restricted far more speech than necessary to prevent true voter deception. It therefore was unlikely to withstand judicial scrutiny.
The better, constitutional way to fight falsehoods that arise during campaigns is to let candidates fight speech with more speech. If an ad is misleading or outright wrong, candidates can and should point it out. Should any depictions of candidates rise to the level of being actually defamatory, Virginia already has laws to address it. Otherwise, the First Amendment protects our right to use expressive tools like AI to enhance political communication.
Our system of government hinges on the freedom to freely express our opinions about candidates for public office. We commend Youngkin for his veto, which will help preserve the First Amendment rights of Virginians and ensure a vibrant, open political discourse.