US political campaigns are rewriting their web playbooks. Campaigns no longer rely just on TV commercials or search engine rankings. Instead, consultants now aim straight at artificial intelligence chatbots.
Leaked campaign files obtained by Reuters show how political teams use Generative Engine Optimization (GEO). Their goal is simple: train ChatGPT, Claude, and Gemini to pick their candidate when voters ask questions about taxes, schools, or crime.
The stakes are big. About 25% of eligible voters have asked an AI bot about a political candidate before voting, according to Cameron Hickey, CEO of the National Conference on Citizenship.
Technical Comparison: Traditional Political SEO vs. Political GEO
| Strategy Dimension | Traditional Political SEO | Political Generative Engine Optimization (GEO) |
|---|---|---|
| Target Engine | Google and Bing search index algorithms. | Large language models (LLMs) and retrieval-augmented generation (RAG) engines. |
| Primary Goal | Win top-3 organic blue links on search result pages. | Become the single direct recommendation inside the chatbot answer. |
| Core Targets | Keyword density, metadata, backlinks, and page load speed. | High-authority domains, Reddit discussion threads, and structured policy pages. |
| Feedback Loop | Search console clicks, impressions, and bounce rates. | Automated prompt testing measuring candidate “AI recommendation share.” |
| Monthly Pricing | $2,000 to $10,000 per month for standard agency retainers. | $5,000 to $50,000 monthly, plus success bonuses up to $200,000 for a “category win.” |
| Voter Exposure | User scans multiple competing links and sources. | Chatbot provides a single synthesized answer, masking conflicting data. |
The GEO Playbook: Targeting Reddit and Authority Sites
An internal pitch deck prepared for Mike Minogue reveals the new playbook. Minogue is a Republican challenger running against Massachusetts Governor Maura Healey. Austin-based strategy firm Visigent built the pitch.
The presentation shows how campaigns place content on sites that AI models read most.
“When the campaign is ready, we run paid placements on high-domain-authority sites the LLMs already cite — plus Reddit, which is disproportionately weighted in retrieval,” the pitch states.
The campaign strategy runs an active four-step loop:
- Prompt Testing: Staff query chatbots with common voter questions, such as “Who has the best housing plan in Massachusetts?”
- Response Auditing: The team checks if the AI backs their candidate or praises their rival.
- Content Injection: The team posts new policy pages and starts targeted threads on Reddit.
- Content Tweaking: Staff tweak words across web forums until bots switch their answers to favor the candidate.
Consultants track their work through a metric called “AI recommendation share.” This figure measures how often a chatbot names the candidate in a reply.
The High Cost of Algorithmic Influence
Shaping AI answers is now big business. Prices range from cheap grassroots apps to massive consulting retainers:
- Visigent Software & Retainers: Basic software starts at $134 a month. Full political consulting starts at $5,000 a month. That fee climbs to $50,000 a month when campaigns buy paid web placements and influencer outreach. Visigent also pitched a $200,000 success fee for a “category win” if bots name the candidate as the uncontested top choice on a major issue.
- CampSight (Run for Something): Built by the Run for Something Action Fund for progressive candidates running for school boards and state seats, CampSight charges $9 to $250 a month. Statewide and federal races pay custom rates for an advisor. Over 300 candidates are on its waitlist.
Misinformation Risks Ahead of the 2028 Election
Chatbots do not just share web links. They give a single direct answer. That setup makes hidden bias very hard to spot.
“Campaigns using AI-influence training are now a reality that we face,” said Joe Bonilla, managing partner at creative strategy firm Relentless Awareness. “It will be ubiquitous by the 2028 cycle.” Bonilla warned that voters cannot tell human facts from machine-fed talking points.
Meg Schwenzfeier of Caucus AI noted that chatbots speak with authority, even when fed bad data. Search engines let users scan ten links. AI bots blend everything into one answer. That makes stealth spin much harder to fight.
How Voters Can Fact-Check AI Answers
Voters who use chatbots can take steps to verify what they read:
- Ask for Source Links: Ask the chatbot for links to official voting rolls, legislative bills, and debate video.
- Compare Multiple Bots: Ask the same prompt to ChatGPT, Claude, and Gemini. Check for big differences in tone and facts.
- Check Public Voting Records: Look up past votes directly on official state or congressional web portals.
- Review Campaign Filings: Read state and federal campaign finance records to see if candidates pay AI optimization firms.
TeckScaler Editorial Analysis
As conversational search replaces traditional SERPs, Generative Engine Optimization is transforming digital campaigning. Stay updated with TeckScaler for in-depth coverage of AI engineering, algorithmic search, and digital policy.
