In a cartoon featured in a campaign mailer sent out in support of Doni Crawford by an outside group, the traces of artificial intelligence are most obvious in the depiction of arms and hands. (Photo by Maddie Poore)
If you want to get to know D.C. mayoral candidate Vincent Orange a little better, you can give him a call at his campaign phone number.
Why is he running? His neighborly voice will tell you, “I’m running because Washington, D.C., deserves experienced leadership and real results.” What are his top priorities? He responds, “My top priorities start with public safety as the foundation because without safe communities, nothing else works.” Each answer is relatively succinct and gets to the point. And each is powered by AI.
Ahead of the June 16 primary, campaigns throughout the District have largely embraced artificial intelligence, employing it to make custom content, design mailers, do research, generate writing, conduct data analysis, translate campaign materials, and spit out slogans and soundbites.
In the city’s first major election featuring heavy use of the technology, candidates say it has helped them conduct more efficient campaigns. But many also warn of limitations and risks, and the use of AI in the city’s political campaigns is entirely unregulated, raising further questions about ethics, transparency, and trust.
For multiple local candidates, generative AI has enabled the creation of campaign content at a fraction of the cost and time.
Jackie Reyes Yanes, who is running for Ward 1 councilmember, plays a campaign song at events that was produced with the help of the technology. It sets her biography to an anthemic pop song inflected with Latin rhythms and the sounds of a cheering crowd. Mayoral candidate Gary Goodweather says his campaign has employed AI to assist with catchphrases and soundbites. Oye Owolewa, an At-Large candidate, appears to have used AI to enhance a campaign video featuring bobbleheads in his likeness bopping around Malcolm X Park. And Dwight Davis, also running for an At-Large seat, used AI to create a go-go beat that will soon be amplified from an “ad truck” touring the city.
Davis acknowledges that using the tech to create music was a missed opportunity for a local band, though he said he tried to reach out to some and didn’t hear back. AI allowed him to produce what he needed, when he needed it.
“Campaigns sometimes have to move so fast,” Davis says.
Others say they strictly limit AI’s use to do research or interpret data.
Janeese Lewis George’s mayoral campaign occasionally utilizes AI tools when conducting data analysis, using it for things like identifying duplicates or standardizing formatting, but says that the technology “should be used judiciously.”
Her campaign doesn’t allow AI to be involved in decision-making or writing, according to spokesperson Amanda Michelle Gomez, pointing to potential factual errors, biases, privacy concerns, and the energy costs affiliated with the technology. “AI must never be used without a human in the loop,” Gomez said.
Goodweather adds that “responsible use” of the technology can be helpful with voter outreach and organizational efficiency.
The mayoral campaign for Kenyan McDuffie did not respond to multiple requests for comment on its use of and attitudes toward AI.
Michael Murphy, who is challenging Charles Allen to represent Ward 6, says he uses AI solely for policy summaries. “I think it’s been fairly obvious who and how campaigns have used AI, and that’s why I avoided using it to the greatest extent possible,” he says.
Ward 1 candidate Terry Lynch draws on AI for campaign literature and says it’s simply the latest tool in campaigns, like social media and cell phones were at one time. Each tool should be used “to the right degree,” he says. “No replacing human input and contact, but [AI] has a role [if] managed correctly.”
Some candidates, though, are largely avoiding the technology altogether.
Ward 1 candidate Aparna Raj says the only AI tool her campaign uses is the language translator Deepl.
"The public shouldn’t have to wonder whether an image is real or AI-generated, question whether an elected official or a chatbot drafted policy, or fact check AI-generated content from their government,” Raj says. “I want the public to know that they can trust me, which is why our campaign doesn’t use generative AI and why on the Council, I would push to have guardrails and checks around any AI usage."
Moral candidate Hope Solomon adds that she’s “pro-AI” like she’s “pro calculators.”
“They’re useful tools. But nobody wants a calculator running the city,” she says. “Use AI to make a campaign more efficient, sure. Just don’t use it to replace the thinking. If you’re asking a computer what your vision for Washington should be, you’ve skipped the part where you were supposed to have one.”
The use of AI tools has already seeped into scores of campaigns across the country. In a recent survey of American Association of Political Consultants members, 83% said they regularly use AI for their work — an increase of 24 percentage points over the prior year.
It can be a potent tool for persuading voters. A 2025 study found that AI chatbots are often better able to sway opinions than humans. The technology can also be used to tweak messages based on highly specific voter groups, and “it can execute these refinements hundreds if not thousands of times daily,” notes a Brennan Center for Justice brief on the use of AI in political ads.
There are no federal laws explicitly banning the use of AI in political campaigns. The Federal Election Commission, however, approved an interpretive rule in 2024 clarifying that the law already “prohibits fraudulent misrepresentation” of candidates in a damaging light, and that the statute is “technology neutral,” thus applying to AI.
Last year, coincidentally on the same day Ward 8 Councilmember Trayon White posted a video of a realistic-looking AI-generated avatar encouraging voters to support him, Ward 6 Councilmember Charles Allen introduced a campaign finance reform bill that would have required all political ads using generative AI to include disclaimers — and campaigns to refrain from dispensing any such ads within 90 days of an election. It was co-sponsored by Councilmembers Christina Henderson, Brianne Nadeau, and Lewis George. Allen also authored a 2024 bill with similar restrictions on AI (with Councilmembers Brooke Pinto, Matthew Frumin, Robert White, and Nadeau.) But neither piece of legislation has moved forward.
“The mere existence of AI deepfakes is sowing doubt in what people can trust,” says Allen, who is currently seeking re-election (he says his campaign has not used AI tools). He notes that his proposed legislation is not an outright ban on AI. “The problem is when AI is used to generate fake images and videos to deceive voters. These kinds of media should have disclosure language accompanying them or not be permitted. Voters should know what they are seeing or hearing was not actually said by a candidate.”
At-large candidate Kevin Chavous, who uses AI for data analysis and volunteer recruitment in his campaign, shares this concern.
“AI is already part of how campaigns operate, and pretending otherwise doesn’t serve voters,” he says. “But as an attorney, I’m clear-eyed about the risks. Deepfakes, fabricated audio, and AI-generated misinformation pose a real threat to election integrity, and D.C. should not wait for a crisis to act.” Chavous says he’s in favor of disclosure requirements regarding AI-generated campaign content, as well as penalties for those who intend to misinform voters.
And of course, the use of AI isn’t likely to stop with the election. Mayor Muriel Bowser’s administration has already embraced AI’s use across government agencies and established a framework for its use in government operations. In February, Bowser also announced a responsible AI training that is mandatory for all D.C. government employees and contractors, which she says is a first among the country’s big cities.
If elected, aides tell The 51st that Lewis George plans to explore ways that AI could potentially improve access and efficiency for D.C. services.
“AI shouldn’t replace workers; it should help them do more of the work they love and more of the work that improves the lives of Washingtonians,” a spokesperson said. “We are blessed to have many of the leading experts in the use of technology in government and [she will] lean on them to understand the best ways to benefit from AI and other tech tools while understanding the very real limitations and risks so we do no harm.”
Solomon similarly argues that AI should be incorporated to improve efficiency. “Government should use AI the same way people use dishwashers: to handle the chores nobody should be wasting time on.” Among these are issuing permits, cutting down on backlogs and identifying waste and fraud, she says. “Residents shouldn’t need a lobbyist, a lawyer, and three follow-up emails just to get a permit or a response from an agency.”
Goodweather says he’d also draw on AI in his administration to “reduce delays and cut red tape” for constituents, as well as boost public safety by finding crime trends and “improving resource deployment,” among other initiatives.
While they make their closing arguments to D.C. voters via mailers, digital ads, and social media — whether AI generated or not — political analysts continue to debate what constitutes an appropriate use of artificial intelligence in campaigns.
“AI isn’t the problem; it’s making stuff up that’s the problem,” says Peter Loge, who heads the School of Media and Public Affairs at The George Washington University and its Project on Ethics in Political Communication.
The campaign veteran and former congressional staffer acknowledges that AI can amplify a falsehood faster and to a larger audience, but he notes the lying itself is nothing new: “People have been lying about politics since the invention of politics.”
Loge also says regulating AI in political campaigns can be murky, as political speech is generally protected by the First Amendment and the boundaries of what is and isn’t AI-generated are not always clear. He offers the example of a fictitious ad depicting a political opponent in front of the Capitol building with Frankenstein-like lightning bolts.
“If I did that because I'm clever on a computer, is that any better than if I get AI to do it?” Loge asks. “Why do we care about AI rather than [the fact that] I’m making my opponent look like a monster?”
But for Allen, the fact that political campaigns will continue to spread falsehoods about their opponents regardless of the technology used doesn’t mean these ads shouldn’t be regulated. “We can’t just wash our hands of any responsibility,” he argues. “We don’t do that with any other product that could cause harm, why would we not regulate AI the same way?”
And candidates may not always have a say in how AI is employed on their behalf. A group affiliated with sports betting companies has sent out reams of flyers in support of Doni Crawford in the At-large special election; the AI-generated cartoon shows the candidate running a race, cheered on by a group of supporters that include a disembodied six-fingered hand, arms melting into the crowd, and an incorrectly rendered D.C. flag, among other anomalies. (Crawford’s campaign declined to comment about the flyer.)
But as At-Large candidate Davis put it: “AI is here … These evolving technologies are here to stay, and we have to figure out a way to live with them.”
Eliza Tebo is a multimedia journalist reporting for regional and national outlets. She specializes in arts and culture features, with bylines in The Washington Post, NPR, Grammy.com, The 51st, WAMU/DCist, Washington City Paper and Arlington Magazine.
With your help, we pursue stories that hold leaders to account, demystify opaque city and civic processes, and celebrate the idiosyncrasies that make us proud to call D.C. home. Put simply, our mission is to make it easier — and more fun — to live in the District. Our members help keep local news free and independent for all:
