The partisan divide in campaign AI use: Republicans generate more content, Democrats govern it more
Republicans are generating far more AI content. Democrats have far more governance.
Don’t forget to join us tomorrow, June 11th, at 1 pm Eastern for our free webinar digging deeper into the numbers about how voters and campaigns are using AI. I’m also excited to announce that Luis Lozada of Democracy Works will be joining us. He’ll discuss how AI platforms are working to ensure voters receive accurate election information and where there’s still room to improve.
Ahead of the webinar, here are some additional findings that didn’t make it into the first newsletter. The partisan tool split from our campaign AI survey generated a lot of conversation — Claude as the Democrat tool, ChatGPT as the Republican tool, and Grok was nearly exclusive to Republicans. So, I went back into the data to see what else emerged.
📊 Republicans are generating. Democrats are governing.
When you break down the use cases by party, a clear pattern emerges.
Republicans are using AI for content creation at roughly double the rate of Democrats. 61% of Republican campaign professionals use AI for image, video, or audio generation, compared with 25% of Democrats. 68% of Republicans use it for ad creative, compared with 39% of Democrats. 71% use AI to draft emails, fundraising copy, and speeches — compared with 50% of Democrats.
If you want to see more graphs like this, visit this dashboard.
Democrats, meanwhile, are more likely to have formal policies governing how staff can use AI. 32% of Democratic respondents have a formal, defined AI policy. Only 10% of Republicans do. And 52% of Republican campaign professionals have no policy at all, compared to 21% of Democrats.
The one place Democrats lead on use cases: research. 68% of Democratic campaign professionals use AI for research, opposition, and issue briefs, compared with 58% of Republicans. Democrats are also more likely to have built custom or internal AI tools — 29% versus 23%. So it’s not that Republicans are using AI more across the board. It’s specifically the content creation stack where the gap opens up.
The conventional read on this might be that Republicans are the pro-AI party and Democrats the cautious ones. I don’t think that’s what the data shows. Democrats are using AI too — but they’re more worried about how voters will react.
When asked directly how concerned they are about voter backlash from AI use, 61% of Democratic campaign professionals say they are very or somewhat concerned, compared with 52% of Republicans. That 9-point gap is modest in absolute terms, but it tracks almost exactly with the governance gap. The campaigns that are more worried about voter reaction are the ones that have built more infrastructure to manage it. Whether that’s the right response to the risk — or enough of one — is a different question.
My read: the capability-governance gap is always a risk, but it compounds when you’re producing at volume. The campaigns generating the most AI content with the least oversight are the ones most likely to have something go sideways and least likely to have a policy that tells them what to do when it does.
On June 11th at 1 pm ET, I'm hosting a live briefing on AI and the 2026 midterms — pulling from our Rainey Center poll of 1,010 likely voters and
our survey of 68 campaign professionals. It's also a live preview of
what my monthly Briefing Network calls look like.
🤔 The finding that complicates the narrative
Despite generating less AI content and worrying more about voter reaction, Democratic campaign professionals are significantly more convinced of AI’s long-term importance: 86% say it will be essential to staying competitive in the next 1-2 election cycles, compared to 65% of Republicans.
That’s a 21-point gap, and it runs counter to the content creation numbers. Democrats are producing less, governing more, and yet are more bullish on AI’s ultimate significance to campaigns. Republicans are producing more with less oversight and are comparatively less convinced it will be decisive.
One way to read this: Democrats are treating AI as a serious institutional challenge — something that requires governance precisely because it matters. Republicans are treating it more as a tactical tool — useful now, but not necessarily the thing that transforms how campaigns operate.
🛠️ Beyond the big models
One question we didn’t cover in the first piece: what tools are campaigns using beyond ChatGPT and Claude?
29% of respondents use custom or internal AI tools built by their teams or vendors. Not off-the-shelf products — bespoke infrastructure. That group is almost entirely at the executive or director level and splits evenly across party lines. This is the leading edge of where campaigns are headed: not just using the public models, but building their own systems on top of them.
Among the other tools in use: AI-powered research and search tools like Perplexity (29%), image generation tools including Midjourney and DALL·E (26%), and transcription tools like Whisper and Otter.ai (26%). Video tools — Opus Clip, Riverside — are at about 11%. The full stack of a modern AI-forward campaign looks much more like a media company than it did two cycles ago.
One last thing
Republican campaigns are producing more AI content faster, and Democratic campaigns are governing it more carefully. At some point, though, those two facts are going to collide — either in an incident that would have been prevented by a policy that didn’t exist, or in a capability advantage that accrues to whoever figured out how to move fast without breaking things.
We don’t know yet which dynamic wins. But we know which campaigns are better positioned if something goes wrong.
If you haven’t built an AI use policy yet — regardless of which side of the aisle you’re on — the window for doing it proactively is still open. I recently released a checklist for any organization that wants to write a policy and isn’t sure where to start. You can get the AI Policy Checklist here.
If this helped you connect some dots, consider sharing it with someone else trying to make sense of where AI, technology, politics, and public trust are all heading next. Word of mouth is still how most people find me, and I’m grateful for every share.
There something like six states that are using AI to cut Medicare benefits to their citizens!