Political campaigns are increasingly utilizing AI-powered chatbots to engage voters through text messages as the midterm elections approach. These bots, designed to mimic the voice of candidates, are capable of having personalized conversations with numerous voters at once. They not only engage in dialogue but also gather valuable insights on what voters seek from their representatives, allowing campaigns to tailor their messages accordingly.
Aaron Sheeks, the CEO of Akillion, an AI service that enables users to run their own language models, noted that many of his clients are currently pursuing political office. “Our mission is to empower voters by providing agencies and campaigns with a trained AI that can address questions on issues like police reform, education, and tax policy," Sheeks explained.
Opinions on the impact of generative AI in political messaging vary widely. While some industry observers see it as a groundbreaking tool for enhancing voter interaction, others dismiss political texts as an ineffective nuisance. Current trends suggest that Republican campaigns are more rapidly integrating AI technologies compared to their Democratic counterparts.
Eric Wilson, a Republican strategist leading the Center for Campaign Innovation, expressed optimism about AI’s potential to make campaigns more engaging and responsive. He stated that generative AI allows campaigns to maximize their resources, although the initial text usually comes from a human before the AI engages further.
Tom Carroll, CEO of Convos—a platform that harnesses AI for political messaging—highlighted a shift away from lengthy text exchanges. He emphasized that their approach provides quick, effective responses, making it seem like voters are conversing with an exceptionally efficient volunteer.
Since its inception last year, Convos has teamed up with ten political campaigns, with plans to support more than 100 in 2023. Currently, they've made significant progress towards that goal.
Marty Santalucia from Vector Political reinforced the advantages of chatbot interactions, noting high engagement rates among voters. "We’ve facilitated millions of texts this year, resulting in tens of thousands of conversations," he stated, stressing that campaigns can listen to voter concerns at an unprecedented scale.
The realm of political texting surged in 2020 as campaigns sought new ways to engage with constituents amidst declining phone call effectiveness and growing wariness about in-person outreach. Josh Justice, CEO of Peerly, insisted that virtually every House or Senate campaign will leverage text messaging this cycle, as it provides a direct line to voters without the interference of social media algorithms.
Despite the potential benefits, concerns about the ethical implications of generative AI in campaigning have surfaced. Justice argues that while AI can enhance data analysis or assist volunteers, it should be clearly disclosed to voters when they are communicating with a bot. "I believe it is unethical to communicate with voters using generative AI without transparency," he declared. The legal landscape surrounding AI is still evolving, with various states requiring candidates to identify when they use virtual assistants.
Nathan Rifkin, co-CEO of Scale to Win, expressed his concerns about the inherent risks of AI, citing the possibility of bots disseminating inaccurate or harmful information. He emphasized the importance of accountability in these interactions.
Many companies providing generative AI solutions for political messaging prefer to keep their use of the technology private, often due to competitive reasons or uncertainty about public reception. Surveys indicate that Democratic candidates tend to be more cautious about embracing AI due to skepticism about government regulation.
Wilson remarked that differing attitudes toward AI in political spheres may reflect broader ideological divides, particularly regarding the technology's environmental impact and labor consequences, which are typically more concerning for Democratic platforms.
Personal experiences shared by voters reflect growing annoyance with political texts. Stefanie Party, a Cleveland resident frustrated by excessive and vague messages, expressed her skepticism about the authenticity of the interactions. "Even if the AI claims to offer personalized information, it feels very impersonal," she shared.
Jessica Alter, co-founder of Tech for Campaigns—a nonprofit aimed at helping Democrats adapt to modern digital strategies—argued that while text messaging remains a viable means for increasing voter engagement, AI should be used as a tool for exploring innovative outreach methods rather than salvaging channels that have lost their effectiveness. "AI is best utilized to discover new approaches rather than to inject life into existing methods that voters already dislike," she concluded.
