Political Campaigns Deploy Generative Engine Optimization to Influence AI Chatbot Answers for Voters

As artificial intelligence becomes a primary information source for the American electorate, political campaigns have begun engineering their platforms to specifically appeal to the Large Language Models that power popular chatbots. Recent data indicates that approximately 25 percent of eligible voters have consulted a chatbot when deciding how to vote, a shift that has forced strategists to move beyond traditional search engine optimization and toward a new frontier known as Generative Engine Optimization, or GEO.

This emerging field focuses on ensuring that when a voter asks a chatbot who the best candidate is for a specific issue—such as housing, taxes, or public safety—the AI returns a specific, curated recommendation. Documents detailing a strategy for Mike Minogue, a Republican challenger to Massachusetts Governor Maura Healey, illustrate the depth of these efforts. The plan sought to influence the outputs of major platforms including ChatGPT, Claude, and Gemini, aiming to position the candidate as the preferred choice across key policy areas.

The tactics involve a multi-pronged approach to feed AI models data that shapes how candidates are described and compared. Strategists propose creating hundreds of AI-friendly policy pages and citations across high-authority websites that chatbots are known to cite frequently. Because AI models often prioritize certain platforms for information retrieval, campaigns are also focusing on producing content for Reddit and YouTube, which carry disproportionate weight in how some models gather data for their responses.

Thomas Howell, CEO of the consulting firm Visigent, noted that campaigns must now compete to become the recommendation inside the AI answer. He suggested that failing to populate the digital environment with specific, optimized information leaves a candidate’s reputation to the whims of whatever data the AI happens to grab. Visigent currently manages six political campaigns and provides software and editorial services designed to monitor and alter how chatbots respond to political queries.

The financial stakes of this digital arms race are significant. Fees for AI monitoring and influence can range from a few hundred dollars a month for basic software access to as much as $50,000 per month for full-service consulting that includes paid media placements on high-domain websites. Some contracts even include a ‘success fee’ of up to $200,000 for achieving a ‘category win,’ which is defined by how often a chatbot returns the candidate as a top recommendation in response to issue-based questions.

While the technology offers new ways for candidates to reach voters, experts have expressed concern regarding the authoritative tone chatbots use when presenting potentially unvetted information. Meg Schwenzfeier of Caucus AI noted that while many current efforts focus on ensuring candidate websites simply appear in results, the risk of misinformation remains high because chatbots can consume vast amounts of data without the traditional filters of human journalism. Joe Bonilla, managing partner of the strategy firm Relentless Awareness, predicted that these techniques will be ubiquitous by the 2028 presidential cycle, warning that the public may soon struggle to distinguish between human-originated content and AI-trained campaign messaging.

On the progressive side, tools like CampSight have emerged to help younger candidates running for local offices monitor what AI models are saying about them. Operated by the Run for Something Action Fund, the tool provides tips on how to influence AI narratives for a monthly fee. Jordan Haines, the organization’s Chief Technology Officer, observed that information posted on platforms like Reddit can appear in AI answers as quickly as the following day, though he noted that some models have shown an increasing ability to discern reputable sources from non-reputable ones.

As the 2026 midterm elections approach, the focus remains on controlling the ‘AI recommendation share.’ Strategists emphasize that this is not merely about visibility, but about shaping the entire information environment that voters now use to navigate complex political landscapes. With more than 300 candidates and organizations already on waiting lists for AI-influence tools, the transition from traditional web searching to AI-driven inquiry appears to be fundamentally altering the mechanics of political persuasion.