As voters face an increasingly complex landscape of candidates and local contests, some are turning to artificial intelligence (AI) tools to help guide their decisions in recent elections. AI chatbots like Claude, ChatGPT, and Google’s Gemini have become a novel resource for individuals seeking faster and more accessible political information, although experts caution about the technology’s limitations.

The emergence of AI assistance in voting is apparent in diverse cases across the country. Mia Taylor, a liberal Democrat from Los Angeles County, used the chatbot Claude to navigate her local election ballot, which featured numerous races. After an initial refusal to provide direct endorsements, the AI offered links to progressive voter guides and detailed explanations of the candidates and issues. Claude advised Taylor to support the incumbent Los Angeles mayor, Karen Bass, over other contenders, a recommendation aligned with her political preferences.

Similarly, in Georgia, Chris Johnson, a self-described libertarian registered Republican who has voted for decades, sought guidance from ChatGPT before the state primary. Johnson asked the AI to identify the most libertarian candidate based on their voting histories. The chatbot suggested Secretary of State Brad Rafensperger, though Johnson later expressed some unease about relying on AI for such decisions, acknowledging the ease but questioning the accuracy of the information.

Users appreciate AI’s ability to condense complex political data into clear, conversational summaries, often finding it more approachable than traditional research methods. Scholars like David G. Rand, a professor at Cornell University who studies AI in political contexts, note that these models are persuasive because they generate plausible facts and clear explanations. Rand himself used AI to analyze a school board candidate forum, finding the tool helpful in forming his voting decisions.

However, concerns about AI’s reliability persist. The output may reflect users’ biases by framing information through their perspectives instead of purely objective data. The parent company of Claude, Anthropic, states its chatbot is designed to approach political questions with balanced, comprehensive, and factual responses, aiming to help users make their own informed choices.

Some voters, such as Jeremiah Hain, a psychotherapist in Los Angeles, have embraced AI’s convenience, using it to quickly assess candidates for mayor and other local offices. Yet experts warn that the perceived authority of AI responses might discourage users from verifying facts independently.

Political scientists urge caution about the sources AI tools draw from, as most current chatbots collect information from across the internet without a curated or verified database of political platforms. Yamil Velez, a Columbia University political science professor, stresses the need for verified data to improve reliability but acknowledges that AI tools are becoming more accurate and remain more accessible than the deep research required for many ballots.

Campaign strategists have taken note of AI’s growing role and are adapting, optimizing online content to be more easily processed by chatbots, potentially influencing AI-generated voter guidance.

Despite these challenges, voters report that AI assistance has increased their confidence in making choices. Robert Siebelink, a Democrat in California, used Claude to narrow down his selection in a gubernatorial race with more than 60 candidates. Similarly, Rikki Powers from Baltimore said AI reduced his research time from about 20 hours to just one, though he remains cautious about sharing his actual votes with AI systems.

As AI technology continues to advance, it may play a more significant role in future elections, offering convenience and clarity while underscoring the ongoing need for vigilance regarding accuracy and bias.