A New Survey Reveals How Campaign Pros Are Actually Using AI

Artificial intelligence use among campaign professionals is booming. A new survey breaks down how they’re actually using the technology.

The survey from a group of newsletter authors and consultants – Anchor Change’s Katie Harbath, Campaign Trend’s Eric Wilson, Chaotic Era’s Kyle Tharp and Doomscroll’s Amanda Elliott – sheds light on how the AI era is actually playing out behind the scenes in politics. It found that political practitioners are racing to deploy AI tools in a bid to stay in the game; 74 percent of respondents say AI use will be essential to compete in the next cycle or two. 

All told, 68 campaign professionals took the online survey, with roughly equal shares of Republicans and Democrats weighing in. 

Here are some key takeaways:

AI Use is the Norm, Not the Exception

Eighty-seven percent of campaign professors say they’re using AI daily, or at least several times a week, according to the survey. 

But so far, AI adoption is overwhelmingly internal. Seventy-seven percent of respondents say they’re using the technology to increase production and about two-thirds say they’re using AI tools for research and news monitoring. 

Of course, some consultants are deploying AI technology for more public-facing tasks. Fifty-nine percent say they use AI to draft copy, while 53 percent say they use it to generate ad creative. Only 23 percent of respondents say that they use AI tools for voter targeting.

“We hear so much externally about deep fakes and AI in ads – all of that. But what we saw from the survey is they’re mainly using it for behind the scenes stuff,” Harbath told Campaigns & Elections. “It’s the communications stuff, the productivity stuff, the research.”

“Most people just kind of ignore that infrastructure part unless they’re on the campaign,” she added. “It’s not as sexy, but that seems to be where the bulk of it is going.”

Many Campaign Professionals Don’t Have a Plan

Despite the boom in AI use among political pros, many still lack a defined policy for how and when to use the technology. 

The survey found that 33 percent of respondents have no AI policy in place whatsoever; no standards or rules outlining what tools their campaigns or organizations can use, what data can and can’t be used and what approval needs to happen.

In a way, Harbath said, that finding isn’t all that surprising. Campaigns are often cobbled together quickly and shut down even faster once a race is over. Many operations – especially smaller ones – lack clear policies on a lot of fronts. 

But public distrust and concerns about AI make it particularly important for campaigns to establish clear guidelines for AI use, Harbath said.

“I feel like this is one area where, because of the sentiment around AI right now, that it’s going to be really important to have policies for how your employees use it and when and how you disclose that to voters,” she said.

There’s a Partisan Divide When it Comes to AI Models

When it comes to AI tools, Democrats are overwhelmingly drawn to Anthropic’s Claude, with 82 percent saying that it’s their model of choice, according to the survey. 

Republicans, on the other hand, appear to favor OpenAI’s ChatGPT; 74 percent of GOP political practitioners said it’s their most used AI tool. Then there’s Grok, which is used by about 45 percent of Republicans and just 4 percent of Democrats.

The exact reasons for the partisan split are unclear, but one factor, according to Harbath, may be Anthropic’s ongoing clash with Secretary of Defense Pete Hegseth and the Pentagon, which deemed the company an “unacceptable risk” to national security earlier this year after an earlier fight over the use of AI in classified systems.

At the same time, Anthropic has sought to bill itself as an advocate for AI safety, which may align more with many Democrats’ more cautious approach to AI adoption. 

Campaign Pros Are Concerned About How Voters Perceive AI Use

More than half of respondents – 59 percent – say they’re at least somewhat concerned about voter backlash to the use of AI in campaigns, the survey found.

That doesn’t mean they’re always telling voters when they’re using it.

Only 12 percent of respondents say they always disclose when AI was deployed for a voter-facing purpose, while 31 percent say they never disclose AI use or haven’t decided on whether to disclose AI use.

Those findings put many campaign professionals at odds with voters. A recent Rainey Center poll found that 78 percent of voters say it’s very or somewhat important for campaigns to disclose when content is AI-generated. That same poll also found deep unease among voters about the use of AI in politics more broadly.

That unease may be part of why many political professionals are reluctant to disclose AI use, Harbath said.

“My gut, at the moment, says that campaigns are concerned about it, but that’s also probably why they’re more reluctant to even disclose it,” Harbath said. “Because if people see that there’s AI usage, then it may make them feel differently about the message and the candidate and stuff like that.”