I confessed to Hutchinson that if I were a politician, I’d be scared to use BattlegroundAI. Generative AI tools are known to “hallucinate,” which is a polite way of saying that the AI sometimes makes up complete fabrications. (In academic jargon, the AI makes stuff up.) I asked how BattlegroundAI ensures that the political content it generates is accurate.
“Nothing is automated,” she replies. Hutchinson points out that the copy of Battleground AI is just a starting point, and that humans on the campaign are supposed to review and approve it. “We might not have a lot of time or a big team, but we definitely review it.”
Of course, there’s a growing movement against the way AI companies train their products on art, writing, and other creative work without asking permission. I asked Hutchinson what she would say to people who might object to the way tools like ChatGPT are trained. “Those are very legitimate concerns,” she says. “We need to talk to Congress. We need to talk to our elected officials.”
“That’s always a possibility,” she replied, when asked if BattlegroundAI would consider offering language models trained exclusively on public domain or licensed data. “We also need to provide the best tools available to people, especially those who are in time-constrained and resource-limited environments. We want to provide consistent results and high-quality information to our users, so the more models we have available, the better for everyone.”
So how does Hutchinson respond to the opposition of progressives, typically aligned with the labor movement, to the automation of advertising copywriting? “It’s clearly a legitimate concern,” she says. “There’s always fear when a new technology emerges. We’re afraid of computers and we’re afraid of light bulbs.”
Hutchinson said she sees it as a way to reduce repetitive tasks, not a replacement for human labor. “I’ve worked in advertising for a long time, and there’s so much repetitive stuff in advertising that, frankly, kills creativity,” she said. “AI takes the tedium out of it.” She sees BattlegroundAI as a lifeline for overworked, underfunded teams.
Taylor Coutts, a Kentucky political strategist who recently started using the service, calls it “very sophisticated,” saying it helps him identify target groups of voters and tailor his messaging to reach them in ways that smaller campaigns have trouble doing. Budgets are tight in gerrymandered battleground races where progressive candidates are vastly outnumbered. “We don’t have millions of dollars,” he says. “Any opportunity we have to be more efficient, we’re looking for it.”
Will voters care if the copy in the digital political ads they see was generated with the help of AI? “I don’t see how it’s any more unethical for an AI to generate content than it is for anonymous staffers or interns to generate content,” says Peter Logue, an associate professor and program director at George Washington University who founded the Project on Ethics in Political Communications.
“If we can mandate the publication of all political writing written with the help of AI, then logically we would have to mandate the publication of all political writing not written by the candidates (emails, advertisements, op-eds, etc.),” he added.
Still, Rosi is concerned about how AI will affect public trust on a macro level, and how it will affect how people respond to political messaging going forward. “One of the risks with AI is not what the technology does, but how people feel about what it delivers,” he says. “For as long as politics has existed, people have been faking images and fabricating facts. The recent focus on generative AI has only heightened people’s already incredibly high levels of skepticism and distrust. If everything can be fake, then maybe nothing is real.”
Hutchinson, meanwhile, is focused on her company’s short-term impact: “We really want to help people right now,” she said. “We’re trying to move as quickly as possible.”
