AI-Powered Political Ads: BattlegroundAI’s Mission to Amplify Progressive Voices and Challenge the MAGA Movement

To review this article again, go to My Profile and then select Saved Stories.

"In Just Minutes, Hundreds of Ads: Startup Believes AI Might Overwhelm MAGA Movement

Tales of AI-created political narratives often resemble those of inebriated individuals igniting fireworks: they're likely to culminate in chaos. WIRED is monitoring how AI is being employed in political campaigns globally, with instances ranging from explicit deepfakes to chatbots that distribute false information. The situation has escalated to a level that has led the US Federal Communications Commission to consider enforcing compulsory disclosures for AI applications in TV and radio commercials."

Amid apprehensions, numerous American electoral efforts are adopting artificial intelligence technologies for content creation. This election season is witnessing an increase in AI-created political materials that are not widely recognized, thanks to companies like BattlegroundAI from Denver. This company specializes in utilizing artificial intelligence to quickly produce digital ad text, boasting on its website the capability to generate "Hundreds of ads in minutes."

BattlegroundAI markets itself as an exclusive resource for progressive movements, explicitly excluding MAGA supporters. The platform has swiftly progressed, unveiling a private beta just six weeks prior and a public beta the previous week. Co-founder and CEO Maya Hutchinson is actively engaging at the Democratic National Convention to draw in additional clientele. According to Hutchinson, the company currently serves about 60 clients. (The platform operates on a freemium basis, offering an enhanced subscription for $19 monthly.)

Maya Hutchinson, who helped establish BattlegroundAI, currently serves as its Chief Executive Officer.

"Hutchinson, a marketing professional whose career kicked off with President Obama's digital reelection team, likens it to adding an additional intern to your group," she shares with WIRED. Our conversation takes place at a picnic table within the McCormick Place Convention Center in Chicago, where she speaks louder to overcome the loud music emanating from a close speaker. "Whether it's managing Facebook or Google advertisements, or crafting scripts for YouTube, we assist in executing these tasks methodically."

BattlegroundAI's platform allows users to choose from among five well-known extensive language models, such as ChatGPT, Claude, and Anthropic, for response generation. The system also enables users to tailor their outputs by adjusting the tone, setting the "creativity level," and specifying the number of different responses they desire for a given prompt. Additionally, it provides advice on targeting specific groups and assists in creating tailored messages for diverse predefined topics, including infrastructure, women’s health, and public safety.

Thanks to BattlegroundAI

BattlegroundAI refused to share real examples of political advertisements made with their tools. Nonetheless, WIRED took the initiative to evaluate the service by setting up a targeted campaign for highly liberal individuals between 88 to 99 years old, focusing on the topic of press liberty. One of the prototype ads opened with, “Don’t let misinformation deceive you behind your reading glasses!”

BattlegroundAI specializes exclusively in creating text content, without providing any AI-generated images or sounds. The firm complies with multiple rules governing the deployment of AI technology in political advertising.

Andy Barr, the managing director at Uplift, a digital advertising agency with Democratic leanings, explains that Battleground is ideally designed for political endeavors because it's developed with political regulations in mind. According to Barr, Uplift has been experimenting with the BattlegroundAI beta version for several weeks now. He mentions that it has been beneficial in generating ideas. While the agency has not yet publicized any advertisements created with Battleground's help, it has employed the platform to formulate concepts, Barr notes.

I admitted to Hutchinson that, were I in the shoes of a politician, the thought of deploying BattlegroundAI would fill me with trepidation. The tendency of generative AI systems to fabricate information—referred to euphemistically as "hallucinating" (or, in more scholarly terms, "bullshitting")—is well-documented. I questioned her about the measures being taken to guarantee the veracity of the political material produced by BattlegroundAI.

"She responds by stating that there is no automation involved. Hutchinson explains that the content produced by BattlegroundAI serves as an initial draft, which is then supposed to be examined and sanctioned by personnel from the campaigns. “Even though you may be pressed for time or lack a large workforce, you are certainly going through it,” she adds.

Certainly, there's growing opposition against the methods AI firms use to educate their systems using art, literature, and various creative outputs without obtaining consent. When I inquire with Hutchinson about her message to individuals who may contest the training practices of platforms like ChatGPT, she acknowledges, "Those concerns are absolutely justified," adding, "It's crucial we engage in discussions with Congress and our representatives."

Inquiring about whether BattlegroundAI is considering providing language models trained exclusively on public domain or licensed data, she responds affirmatively, expressing openness to the idea. She adds that it's crucial to equip individuals, particularly those facing tight schedules or operating in settings with limited resources, with the most effective tools at their disposal. The goal is to ensure reliable outcomes and superior information for users, thereby suggesting that having a wider array of models would be advantageous for all parties involved.

When asked about the concerns from the progressive movement, which typically supports the labor movement, regarding the automation of ad copywriting, Hutchinson acknowledged these worries. "Clearly legitimate fears," she mentioned. "It's the usual apprehension that accompanies the introduction of any novel technology—be it the computer or the light bulb."

Hutchinson presents her viewpoint: In her eyes, it's not about substituting human effort but rather minimizing monotonous tasks. "Having spent a lengthy period in advertising, I can attest to the numerous aspects of it that are redundant and quite frankly, sap one's creative energy," she remarks. "AI eliminates the mundane parts." She envisions BattlegroundAI as a supportive ally for teams that are stretched too thin and lack adequate resources.

Kentucky political strategist Taylor Coots, who has started utilizing the service, praises its advanced features. He notes that it assists in pinpointing specific voter groups and crafting messages that effectively target them, a task that poses challenges for smaller campaigns. In highly competitive races within districts that have been strategically redrawn, where funds are limited and progressive candidates face significant disadvantages, the importance of cost-effective strategies is paramount. Coots emphasizes, "We're not working with multimillion-dollar budgets. We're constantly on the lookout for ways to maximize our resources."

Do voters mind if artificial intelligence assists in creating the text for online political advertisements? Peter Loge, an associate professor and director of a political communication ethics project at George Washington University, questions whether using AI to produce content is any less ethical than letting anonymous employees or interns do it.

He suggests that if there were rules requiring the disclosure of political materials produced with AI assistance, it would naturally follow that any political content—like emails, advertisements, and opinion pieces—not directly created by the candidate should also be made transparent.

Nonetheless, Loge is wary of AI's broader effects on societal confidence, especially regarding its influence on public reactions to political communication. "The issue with AI isn't so much its capabilities, but rather the public's perception of those capabilities," he notes. "The act of fabricating images and concocting false narratives is as old as politics itself. The surge in focus on generative AI has only heightened the already significant skepticism and mistrust among the populace. When the potential for falsehood pervades everything, it leads to a pervasive doubt about the truth."

Hutchinson, on the other hand, is concentrating on the immediate effects her company can have. "Our goal is to assist individuals at this moment," she states. "We're making efforts to accelerate our pace."

Recommended for You…

Direct to your email: A selection of our top stories, curated daily just for you

A Faulty CrowdStrike Patch Disrupted Global Computer Systems

The Major Headline: When Could the Atlantic Ocean Possibly Split?

Introducing the age of excessive online consumption

Additional Content from WIRED

Critiques and Manuals

© 2024 Condé Nast. All rights reserved. Purchases made through our website may result in WIRED receiving a share of the sale, as part of our Affiliate Partnerships with various retailers. Any content found on this website is prohibited from being copied, shared, transmitted, or utilized in any form without explicit written consent from Condé Nast. Ad Choices

Choose a global website