Welcome to the forefront of conversational AI as we explore the fascinating world of AI chatbots in our dedicated blog series. Discover the latest advancements, applications, and strategies that propel the evolution of chatbot technology. From enhancing customer interactions to streamlining business processes, these articles delve into the innovative ways artificial intelligence is shaping the landscape of automated conversational agents. Whether you’re a business owner, developer, or simply intrigued by the future of interactive technology, join us on this journey to unravel the transformative power and endless possibilities of AI chatbots.
JournalismPakistan.com | Published: 26 October 2026 | JP Global Monitoring
WASHINGTON — Political consultants in the United States are developing strategies to influence how AI chatbots describe political candidates and their policies, according to documents Reuters obtained. The emerging approach seeks to shape the information voters receive from services such as ChatGPT, Claude and Google Gemini ahead of the November 3 midterm elections.
A Reuters investigation published on October 26, 2026, detailed a proposal prepared for Mike Minogue, the Republican challenger to Massachusetts Governor Maura Healey. The strategy aimed to improve his standing in chatbot responses to questions about taxes, housing, public safety and other election issues. The documents describe a proposed campaign rather than establishing that the recommendations were implemented or successfully influenced AI responses. Reuters said Minogue's campaign did not respond to requests for comment.
The approach, known as generative engine optimization (GEO), represents a shift from conventional search engine optimization, which focuses on improving a website's position in search results. Instead, GEO seeks to influence the information AI systems retrieve, summarize, and present when users ask questions.
The proposal recommended creating hundreds of policy pages designed for AI discovery, securing paid placements on websites frequently cited by chatbots, and publishing material on Reddit and YouTube. Consultants would then repeatedly test chatbot answers and adjust the online content to improve how the candidate appeared in responses. Visigent, the Austin-based company behind the proposal, offers these services as part of its political consulting business.
Visigent chief executive Thomas Howell told Reuters that the company has six political campaigns among its clients, although he declined to identify them. Political consulting packages begin at approximately $5,000 a month and can reach $50,000 when paid media placements, influencer outreach and related services are included.
The proposal also outlined a potential success fee of $200,000 for achieving a designated category result. The business model reflects a growing interest in measuring how frequently a candidate appears favorably in chatbot recommendations rather than relying exclusively on website traffic, advertising impressions or search rankings.
The potential influence of these services depends partly on how extensively voters use chatbots. Cameron Hickey, chief executive of the National Conference on Citizenship, told Reuters that approximately 25 percent of eligible voters had asked a chatbot about political candidates while deciding how to vote.
A more narrowly focused survey published by Pew Research Center on October 23 provides a different measure. Its survey of 5,273 American adults, conducted October 12-18, found that 8 percent were using AI chatbots to obtain information about the 2026 elections. Among those users, 76 percent sought information about candidates' positions, 71 percent checked the accuracy of political claims encountered elsewhere, and 40 percent sought suggestions about voting choices. The Reuters-cited estimate and Pew's findings use different definitions and populations and should not be treated as directly comparable.
The consulting market is not confined to Republican campaigns. Reuters also examined CampSight, a service operated by the progressive organization Run for Something Action Fund. CampSight offers monitoring and guidance on chatbot responses, with standard services priced between $9 and $250 monthly and customized arrangements for larger campaigns. Its operators reported observing information from Reddit appearing in subsequent chatbot responses, although they also said AI systems had generally proved more capable of distinguishing reputable information than initially feared.
The growing interest in influencing chatbot answers presents challenges for AI developers attempting to provide accurate and politically balanced information. In an October 14 announcement, Anthropic said an analysis of approximately 250,000 Claude conversations found that 0.51 percent specifically concerned elections or voting. Requests about whom or what to vote for accounted for approximately 0.01 percent of the conversations examined.
Anthropic outlined measures to connect users with official voting information and strengthen protections against deceptive political activity. OpenAI, in an October 20 update to its election safeguards, said it had introduced voting information from Democracy Works into ChatGPT and maintained restrictions on using its products to generate or distribute large-scale political campaign messaging.
Those safeguards do not necessarily address every aspect of the strategy described by Reuters. There is an important distinction between using an AI service to create prohibited political content and publishing campaign material on external websites in the hope that an AI system will later retrieve it. The latter raises questions about how chatbots assess source credibility, recognize paid material and distinguish campaign advocacy from independent information.
Research has already identified weaknesses in the way chatbots handle election-related information. In September, the Institute for Strategic Dialogue published an assessment involving 2,400 responses from six AI models, examining accuracy, source reliability, responses to disputed claims, and differences between languages. The research highlighted the importance of reliable sources when automated systems provide information about voting procedures and contested political issues.
For news organizations, the development adds a new dimension to election coverage. Traditional reporting examines campaign advertisements, financing, political statements, and efforts to influence search and social media audiences. AI-generated responses create another point of scrutiny because readers may encounter political claims through a synthesized answer without immediately recognizing the original publisher or the interests behind the information.
The implications extend beyond American elections. In countries such as Pakistan, where political debate involves competing narratives across television, digital news outlets, and social platforms, similar techniques could complicate efforts to distinguish independently verified reporting from campaign-produced material. This is a potential risk rather than evidence that the specific practices identified by Reuters are being used in Pakistan.
The immediate question is not whether political campaigns can make their policies accessible to AI systems, but whether promotional spending and coordinated publishing can distort the information presented to voters. As political consultants compete to influence chatbot responses, the credibility of the underlying sources and the transparency of AI-generated answers will become increasingly important to journalists, election observers and the public.
ATTRIBUTION: Based primarily on Reuters reporting by Mike Stone published October 26, 2026. Additional context comes from Pew Research Center (October 23, 2026), Anthropic (October 14, 2026), OpenAI (October 20, 2026), and the Institute for Strategic Dialogue (September 3, 2026).
PHOTO: AI-generated; for illustrative purposes only.
GEO is an approach aimed at influencing what generative AI systems retrieve and present by tailoring online content and sources they rely on.
Campaigns would publish targeted policy pages, buy placements on sites often cited by chatbots, post on platforms like Reddit and YouTube, and repeatedly test chatbot responses.
Proposals reference services such as ChatGPT, Google Gemini and Claude, and target sites and social platforms that chatbots commonly cite.
No. The documents describe recommended strategies but do not establish that campaigns implemented them or successfully altered chatbot responses.
Dig deeper, ask anything — get instant context, background, and clarity.
Not sure what to choose? Try one of these.
October 19, 2026: India's Supreme Court held internet access is a fundamental right under Art. 21 and 19(1)(a), rejecting blanket bans on children but allowing targeted measures.
October 15, 2026: A Citizen Lab investigation found online ads bought by Israeli firm Discover-Tech delivered Candiru spyware to Al Jazeera devices using a clickless targeted-ad technique.
October 15, 2026: Ofcom is advancing a probe into whether X met UK online safety duties, focusing on illegal content and complaints, and saying Grok has tightened explicit content rules.
October 14, 2026: Mexico ordered digital platforms to strengthen safeguards for children and teens against harmful content and to report within 30 days measures implemented.
October 13, 2026: Meta has asked the High Court to block Ofcom's use of worldwide revenue and group liability to set online safety penalty caps, arguing fines should be proportionate and tied to UK-linked activity.
October 09, 2026: Philippine Senate passed a bill barring under-18s from accounts on high-risk social media platforms, mandating age checks and imposing hefty fines.
October 07, 2026: Türkiye restricted 1,270 X accounts in September; about 90% were unavailable locally and most orders lack a clear legal basis, affecting media and journalists.
September 30, 2026: Google has sued an EU court to block two DMA orders that would force it to share Android features and Search data, saying they damage user privacy and security.
October 26, 2026 US political consultants are using generative engine optimization tactics to shape how AI chatbots portray candidates and policies to influence what voters read before the midterm elections.
October 26, 2026 Pakistani journalists facing violent threats can combine safety planning, evidence preservation and legal advice, without assuming any route guarantees safety.
October 26, 2026 Dabanga reports Sudanese journalist Rasha Hassan was detained over a school-fees report repost, prompting condemnation from the journalists’ syndicate.
October 26, 2026 Bangladesh preacher Enayetullah Abbasi faces a legal notice demanding an apology within seven days over remarks allegedly urging violence against journalists.
October 25, 2026 Assaults, detentions, legal and financial pressures, and protests this week weakened independent journalism across many countries, despite a few limited legal reprieves.
Sep 27, 2026
Sep 26, 2026
Oct 04, 2026
Sep 27, 2026
Sep 27, 2026
Oct 11, 2026
Oct 12, 2026
Oct 18, 2026
Sep 30, 2026
Oct 17, 2026
Founded in 2009 by journalists Stephen Manuel and Imran Naeem Ahmad
Disclaimer: The news content does not necessarily reflect the opinion and views of JournalismPakistan.com
This publication uses AI-assisted tools under human editorial oversight.
Please wait while we prepare your result…
