Americans are growing skeptical of all things AI. A recent poll from Emerson College found that nearly twice as many people believe AI will have more of a negative impact on the economy than a positive one, and 46% think the technology will have a negative impact on the environment compared to just 21% who believe it will have a positive impact. Morning Consult has found a growing number of people supporting bans on constructing data centers. As the public sours on AI, politicians are pushing to position themselves as the anti-AI party ahead of the midterms. Some are having more success than others.

The Democratic Party should be in pole position to be the party that opposes AI exceptionalism. The Trump administration has done everything in its power to back Big Tech, including signing an executive order that aims to ban states from passing AI regulations, and Republican lawmakers have tried to sneak favors to the AI industry into bills all year. Democrats, meanwhile, have led the way in creating protections against AI. California and New York, both led by Democratic lawmakers, have passed meaningful regulations meant to rein in AI companies and create restrictions on how AI models can be deployed for things like algorithmic price fixing.

Those are positions that have broad support. A Gallup poll found that 80% of the people it polled believe the government should regulate AI, even if it means that industry growth will slow. Pew Research found that people are pretty skeptical of that growth leading to anything in the short term, with just 17% of participants expressing the belief that AI will have a positive impact on the country over the next two decades.

And yet, some of the leading voices in the party seem set to cede ground as anti-AI sentiment starts to grow in Republican circles. Business Insider recently published a story highlighting how several Senators have started playing with AI models. Among them is Elizabeth Warren. Earlier this year, Warren said she “might as well go out on the street and ask a random person a question and see what words they come up with” rather than ask a chatbot. Now she says she uses ChatGPT for information gathering and says it can “get an answer that’s better than a straight Google answer,” per Business Insider.

That’s not to say that she’s abandoned the idea of regulating the technology, but it does put her in the camp of politicians who have seemingly decided to surrender the moment, likely in an effort to appear more pragmatic without realizing that it comes off as nihilistic. Democratic Senator Chris Murphy perhaps best encapsulates this shift. He told Business Insider, “I use it, despite the fact that I think it’s going to destroy us.”

Meanwhile, the anti-AI movement in the Republican Party is rising. Florida Governor Ron DeSantis, likely trying to find a way to wedge himself back into the national conversation for a post-Trump party, has positioned himself as the biggest figure to oppose AI on his side of the aisle, proposing several bills to protect Floridians from AI overreach and data center buildouts. Other figures like Marjorie Taylor Greene and Steve Bannon have also taken stances against Big Tech’s AI expansion, though how seriously they are viewed in right-wing circles these days is unclear.

The progressive wing of the Democratic Party seems to be doing its best to anchor itself as the party of the people, not AI or the companies that make it. Bernie Sanders, consistent as ever, has gone the furthest of anyone, proposing a nationwide moratorium on building AI data centers as local communities fight to keep the massive infrastructure from being set up in their backyard. Others have followed his lead. Gretchen Whitmer, the Democratic governor of Michigan, recently called for a ban on data centers in her state.

But whether those voices have enough power to win out over the corporatist wing is certainly an open question. The fact that Warren and Murphy are already changing their tune suggests the party might once again find a way to throw away an opportunity to win.