Sloptimisation for Chatbots as Political Campaign Advertising: How Political Campaigns Are Trying to Shape AI Answers
PoliticalMarketerIn shortSloptimisation for chatbots is the practice of producing, placing, repeating, or structuring political information so that AI assistants are more likely to retrieve it, summarize it, cite it, or use it when answering voter questions.
On this page22 sections
The term is closely related to Generative Engine Optimization (GEO) and Answer Engine Optimization (AEO), but sloptimisation has a more negative meaning. It describes campaigns that optimize primarily for machines through repetitive, self-serving, synthetic, paid, or strategically distributed content. For political campaigns, the goal can shift from winning a search ranking or ad impression to influencing the answer a voter receives when asking an AI assistant about a candidate, policy, issue, or election.
Sloptimisation Is Not the Same as GEO or AEO
Generative Engine Optimization and Answer Engine Optimization are broad practices concerned with how information appears inside AI-generated answers. Sloptimisation is a narrower and more critical term for attempts to game those systems through content created mainly to influence machine retrieval or recommendations.
A legitimate political campaign might publish an accurate policy page explaining a candidate's housing proposal, voting record, biography, or tax position. Making that material clear, current, well sourced, and easy for machines to interpret falls within ordinary information publishing.
Sloptimisation begins when machine influence becomes the primary purpose of the content.
Typical characteristics can include:
- Large volumes of similar candidate or policy pages created mainly for AI retrieval.
- Repeated favorable descriptions across multiple websites.
- Self-serving comparison content that repeatedly places one candidate first.
- Paid placements on websites that AI systems commonly retrieve.
- Artificial discussions or coordinated mentions intended to create apparent consensus.
- Machine-generated articles with little independent value for voters.
- Hidden instructions or manipulated content designed to affect AI recommendations.
- Content distribution aimed at increasing the probability that one political narrative dominates an AI-generated answer.
The distinction matters because GEO itself is not necessarily deceptive. Sloptimisation describes the point where optimization can become information manipulation.
The term gained wider attention in June 2026 after reporting described companies producing machine-targeted material designed to influence chatbot recommendations. Political campaigns soon became a natural extension of the same idea because voters were increasingly using AI assistants to research candidates.
Political Chatbots Existed Long Before Sloptimisation
Political chatbot use did not begin with generative AI. The current practice grew from more than a decade of political bots, messaging automation, voter-information chatbots, conversational campaign tools, and algorithmic political communication.
During the early 2010s, political actors experimented heavily with automated social accounts. Researchers documented bots spreading campaign links, creating artificial conversations, amplifying political messages, and making support for candidates appear larger than it was.
The 2016 U.S. election marked an important change. Chat applications were used for voter registration, election information, political news, campaign communication, and turnout efforts. Messaging services began supporting bots that could answer questions, locate polling places, register voters, or distribute campaign information. A post-election study described 2016 as the first U.S. election cycle in which chat platforms collectively attempted voter registration and political engagement at meaningful scale.
These systems were very different from today's large language models. Most followed scripted decision trees or limited keyword systems. They usually answered from predefined databases rather than generating new political explanations.
In 2017, an AI-based political chatbot called SAM appeared in New Zealand. It was designed to discuss political topics through conversational interaction and was described as an AI political representative rather than a normal campaign messaging bot.
Political chatbot use became more directly connected to campaigning in later elections. Research into Israel's 2019 elections documented a campaign Messenger bot that collected voter information, repeated campaign messaging, and encouraged political action. Researchers later described a cycle involving data collection, repeated messaging, and activation of supporters.
During the 2020 U.S. presidential campaign, conversational AI was also used through messaging services to interact with potential voters. These systems represented a shift from simple automated replies toward more flexible political conversations.
The important distinction is that early political chatbots were normally controlled by campaigns and directly encountered by voters. Sloptimisation reverses part of that relationship. A campaign may never operate the chatbot that reaches the voter. The campaign tries to influence an independent AI assistant by changing the information that assistant finds online.
Generative AI Changed the Target From the Voter to the Information System
Generative AI created a second audience for political communication. Campaign websites are still written for voters, journalists, donors, volunteers, and search engines, but some campaign content is now also written with AI retrieval systems in mind.
Modern AI assistants can search current web information, retrieve passages from multiple sources, compare entities, and produce one synthesized response.
A voter no longer needs to visit ten candidate pages.
The voter can ask an AI assistant:
"Which candidate supports lower property taxes?"
"Compare the candidates on housing."
"What has this candidate said about public safety?"
"Which candidate has the strongest small-business policy?"
The AI system may retrieve several sources, select specific passages, combine them with model knowledge, and generate a single response.
That creates a new political communication objective.
Traditional SEO asks whether the candidate appears in search results.
Political GEO asks whether information about the candidate appears in an AI answer.
Sloptimisation goes further and attempts to increase the probability that the AI's synthesis favors a particular narrative.
The Term Sloptimisation Became Politically Important in 2026
The political meaning of sloptimisation became much clearer during the 2026 U.S. midterm cycle. Reporting during July and October documented campaigns monitoring what AI systems said about candidates and altering their online information in response.
One 2026 state legislative candidate discovered that AI assistants had limited information connecting his background to his small-business position. The candidate added a question-and-answer section to his website. Later tests found that the AI systems connected his biography more clearly with his policy position. The case demonstrated that chatbot answers about lesser-known candidates can change after new campaign information becomes retrievable online.
By October, the practice had become more commercial.
Documents reviewed by reporters described a proposed strategy for a statewide challenger that included hundreds of AI-friendly policy pages, monitoring chatbot responses, content placed across websites already used by AI systems, paid media, Reddit material, YouTube content, and repeated testing of how candidates appeared inside generated answers.
These reports do not prove that every proposed tactic works. They do show that political consultants are selling chatbot visibility as a campaign service.
That difference matters.
Sloptimisation moved from a theory about manipulating commercial recommendations to an emerging political campaign practice within a few months.
Why Campaigns Started Paying Attention to AI Answers
Political campaigns became interested in chatbot visibility because AI has entered the voter information process even though voter trust remains low.
A national U.S. survey released in October 2026 questioned 1,568 likely general-election voters between September 14 and September 22. It found that 60 percent used AI tools at least several times per month and 25 percent used them daily.
Election-specific use was much lower.
About 19 percent reported ever using AI for information about politics, candidates, elections, or voting. About 11 percent had done so during the 2026 cycle, while about 4 percent had done so during the previous month.
Exposure through AI-generated search summaries was much higher. At least 38 percent said they had received election information through an AI-generated search summary, and only around half recognized that the information had been generated by AI.
Trust remained very low. Only 1 percent named AI tools as their most credible source of election information.
These numbers reveal why chatbot optimization matters even without mass intentional use.
A voter does not need to open a dedicated chatbot and explicitly request political advice. AI-generated summaries can appear during ordinary web searches.
Political sloptimisation therefore targets both intentional chatbot research and incidental AI-mediated exposure.
How Political Sloptimisation Can Reach an AI Answer
Political sloptimisation operates through the information chain between online publishing and AI response generation. The campaign does not directly control the final answer. It attempts to affect one or more inputs that the AI system uses.
A simplified process looks like this:
Political information is published.
Candidate biographies, issue pages, interviews, public records, videos, news coverage, voter guides, social posts, discussion forums, government material, and third-party pages enter the public web.
An AI system searches or retrieves information.
For current political questions, many AI assistants retrieve live or recently indexed material rather than relying only on their original training data.
The system selects passages.
Retrieval systems try to identify text that appears relevant to the user's request.
Sources receive different levels of weight.
Authority, relevance, freshness, source type, wording, query match, retrieval architecture, and model design can affect which information is selected.
The model synthesizes an answer.
The AI combines retrieved information into a response, comparison, summary, or sometimes a recommendation-like output.
Sloptimisation tries to influence the stages before synthesis.
The campaign's problem is therefore different from buying a normal political ad. The campaign is trying to change the informational material from which another system produces the message.
Why Repetition Can Matter Without Becoming Truth
Repeated information can increase retrievability without making the underlying statement more accurate. That distinction is central to the risk created by sloptimisation.
AI retrieval systems can encounter similar language on candidate websites, videos, forums, news-like pages, policy summaries, directories, and third-party sites.
Multiple mentions can create several retrieval opportunities.
The model still may reject the material, prefer stronger sources, find conflicting information, or produce a neutral comparison. Current systems have also become better at distinguishing some low-quality sources from stronger ones. A campaign cannot reliably control an AI assistant merely by publishing large quantities of text.
Sloptimisation should therefore not be described as "training the chatbot" in the technical model-training sense.
Most reported political GEO activity is closer to influencing retrieval, source selection, and synthesis.
The underlying model normally is not being retrained on a campaign's new pages every time content changes.
Sloptimisation Is a New Influence Layer, Not Automatically a Political Ad
Sloptimisation does not fit neatly into the traditional definition of political advertising. Some activity resembles ordinary campaign publishing, some resembles public relations, some resembles search optimization, and some can involve paid political promotion.
A conventional digital political advertisement normally has identifiable components.
A campaign or political group pays for distribution. A platform delivers the advertisement to an audience. The campaign controls the creative message. Impressions, reach, frequency, clicks, and conversions can often be measured.
Chatbot influence works differently.
A campaign might publish a policy page without buying distribution. Weeks later, an AI system may retrieve one paragraph and combine it with information from several other sources.
The campaign did not purchase that answer.
The campaign did not fully write that answer.
The campaign may not even know that the answer was generated.
The activity moves closer to advertising when paid media, sponsored publishing, influencer promotion, or purchased third-party placements are used as part of the strategy.
Legal treatment can therefore depend on jurisdiction, sponsorship, coordination, payment, disclosure requirements, and the exact activity involved.
Political campaigns should not assume that content created for AI retrieval falls outside existing campaign-finance, advertising, disclosure, privacy, or election rules merely because the final message appears inside an AI answer.
2026 Campaign Strategies Show How Far the Practice Could Go
Reported 2026 campaign proposals show that political GEO can range from simple information cleanup to large-scale attempts to shape the wider information environment.
At the low end, campaigns can monitor AI descriptions of a candidate and correct missing or outdated public information.
At the aggressive end, reporting described proposals involving hundreds of policy documents, paid placements, social and video content, third-party citations, influencer activity, and repeated monitoring of model outputs. Some consultants were reported to price services according to an "AI recommendation share," meaning how often a candidate appeared in chatbot responses.
Reported pricing ranged widely. Monitoring services could cost relatively little, while comprehensive political consulting involving media placements could reach tens of thousands of dollars per month. One proposal included a much larger performance fee tied to winning a defined recommendation category.
These numbers demonstrate the emergence of a market.
They do not demonstrate that spending more produces predictable changes in voter behavior.
That distinction should remain explicit whenever political GEO performance is discussed.
AI Recommendation Share Is Not the Same as Vote Persuasion
AI recommendation share measures visibility inside generated responses. It does not measure whether the chatbot changed a person's vote.
Several different metrics are needed to assess political chatbot influence responsibly.
Answer presence measures how often a candidate appears in relevant chatbot responses.
Issue association measures whether the candidate is correctly connected with a specific policy or topic.
Factual consistency measures whether generated answers accurately represent the candidate's documented position.
Source quality examines whether answers rely on primary, reputable, current, or questionable material.
Citation presence measures whether generated answers expose sources that users can inspect.
Response volatility measures how much an answer changes across time, prompts, models, locations, or retrieval conditions.
Recommendation frequency measures how often a candidate is placed first, described favorably, or matched with a user's stated criteria.
Persuasion effect requires actual voter research. Visibility alone cannot establish persuasion.
A campaign can improve answer visibility while producing no meaningful change in candidate preference.
Likewise, a chatbot may influence political opinion even when the campaign never optimized for that chatbot.
Political Chatbots Can Persuade Voters, but Effects Depend on Context
Research increasingly shows that conversational AI can affect political attitudes, which raises the stakes around attempts to shape the information those systems receive.
A 2023 experiment involving 360 participants studied political chatbot anthropomorphism and different message appeals. Voting intention varied according to how human-like the chatbot appeared and the type of message it delivered. The interaction also differed according to political ideology.
Larger studies later examined generative AI.
Three preregistered experiments involving 4,829 participants found that LLM-generated political messages changed attitudes toward several public policies compared with neutral messages. The AI-generated messages were broadly comparable with lay-human political messages in persuasive effect.
Research published in 2025 tested interactive AI conversations during the 2024 U.S. presidential election and the 2025 Canadian and Polish elections. The experiments found measurable changes in candidate preferences following conversations with AI systems advocating for candidates. Researchers reported effects larger than those commonly measured for traditional campaign video advertising in comparable experimental settings. They also found that some supporting statements generated during the conversations were inaccurate.
Another experiment involving 739 participants found a more complicated result. Political AI chatbots increased perceived social presence and likability but had difficulty matching traditional campaign video on authenticity.
The research does not show that sloptimisation wins elections.
It shows why control over chatbot information has political value.
Chatbot Recommendations Create a Different Risk From Search Rankings
A ranked search page normally exposes several competing sources. A chatbot often compresses many sources into one fluent answer, which can reduce the visibility of disagreement.
This creates three important effects.
First, synthesis can hide source boundaries. A voter may not know which sentence came from a campaign website, government record, discussion forum, news report, or third-party page.
Second, conversational language can sound authoritative even when the underlying source material is weak.
Third, users can ask follow-up questions. A political information session can become a personalized conversation rather than a single exposure.
A November 2, 2026 test of five major AI assistants illustrated the recommendation problem. The systems initially resisted direct requests for voting recommendations. After follow-up prompts, four of the five produced candidate lists matching specified political preferences in the test. The remaining system continued refusing to recommend candidates.
The test was intentionally designed to pressure the systems and does not describe normal usage by every voter.
It still demonstrates that the boundary between neutral political information and recommendation behavior can be unstable.
The Biggest Risk Is Source Laundering
Source laundering occurs when partisan or self-interested material enters an AI response without its political origin remaining obvious to the voter.
A candidate can legally describe a policy on a campaign website.
Problems become harder to detect when that message appears across supposedly independent pages, synthetic news-style websites, coordinated discussions, paid articles, or undisclosed promotional material.
The chatbot can then repeat the idea in its own neutral-sounding voice.
The voter may experience the result as an independent AI judgment even though political actors helped create the source environment.
This risk becomes larger when several apparently separate sources originate from the same coordinated campaign.
The volume of sources can create the appearance of broad agreement even when the underlying information comes from one political operation.
Sloptimisation Can Reward Candidates With More Digital Material
AI-mediated political information can create an information asymmetry between candidates.
Incumbents often have years of government pages, legislative records, speeches, hearings, official biographies, interviews, news coverage, and public documents available online.
New candidates may have very little machine-readable information.
An AI assistant can therefore produce detailed answers about the incumbent while providing vague descriptions of a challenger.
Political GEO can partly respond to that problem by improving legitimate public information about lesser-known candidates.
Sloptimisation can create the opposite problem.
A well-funded campaign could flood the web with far more machine-oriented material than a smaller campaign can produce.
The result becomes a contest over information volume and distribution rather than simply factual quality.
Responsible Political AI Readiness Is Different From Sloptimisation
A transparent campaign information strategy can make public political information easier for both voters and AI systems to interpret without flooding the web or creating artificial consensus.
Responsible information publishing emphasizes accuracy, provenance, disclosure, and correction.
Candidate websites can clearly identify who operates the site.
Policy pages can describe the candidate's actual position using stable terminology.
Dates can show when material was published or updated.
Primary documents can support statements where available.
Biographical information can remain consistent across public profiles.
Corrections can be published when facts change.
Paid political material can carry the disclosures required by applicable rules.
Machine readability can support human readability rather than replacing it.
The objective is to make accurate campaign information available when an AI system searches for it, not to manufacture apparent third-party agreement.
That line provides a practical distinction between AI information readiness and sloptimisation.
Platform Rules Can Limit How Far Political AI Optimization Goes
AI providers are developing election rules that can restrict direct political campaign uses even when campaign information remains searchable on the open web.
As of November 5, 2026, OpenAI prohibits political advertisements on its platform and prohibits using its services to create or distribute scaled campaign messaging advocating for or against candidates, parties, or ballot measures. OpenAI also states that its services cannot be used to operate public-facing campaign chatbots or automate political outreach. Human-directed campaign work such as internal research, planning, compliance, translation, and staff briefings can remain permitted under the stated rules.
Other AI providers maintain their own election, deceptive-activity, civic-information, and political-content policies.
Rules can change quickly.
A strategy permitted by one model provider may be restricted by another, and a platform policy does not replace election law or campaign-finance law.
Sloptimisation Also Creates a Defensive Problem for Campaigns
Political GEO is not only an offensive visibility issue. Campaigns also face the problem of incorrect, outdated, fabricated, or adversarial material entering chatbot responses.
A campaign may discover that an AI assistant:
- Confuses candidates with similar names.
- Attributes another person's policy position to the candidate.
- Uses an outdated biography.
- Repeats a false rumor.
- Cites low-quality political material.
- Describes a candidate as running in the wrong race.
- Misstates an endorsement.
- Invents a policy position.
- Fails to recognize that a candidate withdrew.
- Produces inconsistent answers across different prompts.
Research around elections has documented fabricated candidates, false scandals, incorrect candidate status, and incorrect policy descriptions in AI-generated political answers.
For that reason, monitoring chatbot accuracy can serve a defensive information-quality purpose even when a campaign has no intention of manipulating recommendations.
AI Political Information Can Also Improve Voter Knowledge
Chatbots are not inherently harmful political information systems. Research has also found cases where source-grounded AI tools helped people understand election information.
A study around the 2024 U.K. general election surveyed 2,499 adults and found that about 13 percent of eligible voters reported using conversational AI to obtain information directly relevant to their electoral choice. Separate randomized experiments involving 2,858 participants found that people researching political issues through conversational AI became more knowledgeable at roughly the same level as participants using traditional search.
A voter-information chatbot built around official California ballot-guide information also increased engagement with election information and improved participants' ability to answer detailed questions about ballot measures. Researchers did not find a clear effect on turnout or the direction of voting.
These findings show why political sloptimisation needs to be separated from ordinary AI-assisted civic information.
The technology can reduce information costs.
The risk appears when actors manipulate the source environment, hide sponsorship, create false consensus, or push unreliable political information into generated answers.
What Political Campaign Advertising Looks Like After Sloptimisation
Political advertising in the chatbot era is expanding beyond impressions, clicks, video views, reach, and frequency. Campaign strategists now have to consider the information systems that stand between political content and voters.
The emerging model has three layers.
The first layer is direct campaign communication, including speeches, websites, advertisements, emails, social posts, videos, and public statements.
The second layer is information retrieval, where search engines and AI systems determine which sources are relevant to a political query.
The third layer is AI synthesis, where an assistant converts selected material into a personalized explanation, comparison, or answer.
Sloptimisation targets the second and third layers.
This makes the AI system an intermediary between campaign communication and voter interpretation.
That intermediary can summarize accurately, omit information, introduce errors, compare candidates, reject weak sources, or unexpectedly repeat partisan material.
Campaign control therefore decreases at the same time that the political importance of the generated answer increases.
The 2028 Election Cycle Will Test Whether Sloptimisation Becomes Standard Campaign Practice
The 2026 cycle established political chatbot optimization as a visible campaign category, but it has not established a reliable formula for controlling AI answers.
Consultants are already selling monitoring, content strategy, placement, and recommendation metrics. Campaigns are experimenting with machine-readable issue material. Researchers are testing chatbot persuasion. AI companies are developing election safeguards. News organizations and civic groups are auditing political answers.
The next major test will be whether these separate activities consolidate into a standard campaign discipline by the 2028 U.S. presidential election.
Several factors will determine what happens next:
- How many voters use AI for candidate research.
- Whether AI-generated search summaries replace more traditional search activity.
- How strongly models weight primary political sources.
- Whether models become better at identifying coordinated or synthetic content.
- Whether political advertisers must disclose content designed to influence AI retrieval.
- Whether campaign-finance rules cover paid GEO placements.
- Whether AI systems clearly expose source provenance.
- Whether independent auditing becomes common.
- Whether voters trust AI-generated political information.
- Whether recommendation behavior becomes stricter or more permissive.
Current reporting shows political consultants expect AI-answer competition to expand by 2028. That expectation remains a forecast, not a demonstrated outcome.
What Sloptimisation Means for Political Campaigns Now
Sloptimisation represents a shift from campaigning only for human attention to competing over the information that machines use to answer human questions.
The practice emerged from commercial attempts to influence AI recommendations, but political campaigns adopted the same logic as voters began asking AI systems about candidates, policies, elections, and voting decisions.
The technology creates legitimate uses. Campaigns need accurate digital records. Candidates need current policy information. AI assistants need reliable sources. Voters can benefit from faster access to complex political information.
The same system also creates incentives for mass content production, paid source placement, synthetic consensus, hidden sponsorship, recommendation manipulation, and misinformation.
That tension defines political sloptimisation.
The central issue is no longer simply whether a campaign ranks first in search or reaches a voter with an advertisement. The new issue is whether political actors can shape the body of information from which AI systems construct answers, while the voter experiences those answers as neutral assistance.
That makes transparency, source quality, factual accuracy, sponsorship disclosure, model neutrality, and independent measurement central to the next phase of political communication.
Sloptimisation for chatbots is becoming a new form of political influence because campaigns are no longer competing only for voter attention, search rankings, or ad impressions. They are also competing to shape the information AI systems retrieve when voters ask about candidates, policies, elections, and public issues.
The practice can range from legitimate efforts to keep candidate information accurate and machine-readable to aggressive strategies involving mass content production, paid placements, coordinated mentions, and repeated attempts to influence chatbot recommendations. The main concern is transparency. Voters may receive a polished AI-generated answer without realizing that some of the information behind it was deliberately engineered by political actors.
The 2026 election cycle showed that political GEO, chatbot monitoring, and AI recommendation tracking are moving from experimental tactics into professional campaign strategy. Yet visibility inside an AI answer does not automatically translate into persuasion, trust, or votes. Campaigns, regulators, researchers, and AI providers still need better measurement, disclosure standards, source verification, and independent auditing.
As AI assistants become a larger part of how voters research elections, political communication will increasingly depend on who supplies the information, how AI systems select it, and whether voters can identify its origin. Sloptimisation therefore raises a broader democratic issue: political campaigns may increasingly compete not only to persuade voters directly, but also to influence the information systems that voters rely on to make political decisions.
Sloptimisation for Chatbots as Political Campaign Advertising: FAQs
Sloptimisation for chatbots is the practice of creating, repeating, structuring, or distributing political content so AI assistants are more likely to retrieve it and use it when answering voter questions. Campaigns may optimize candidate biographies, policy pages, interviews, videos, third-party mentions, and other online material to influence how AI systems describe candidates or issues.
Generative Engine Optimization (GEO) and Answer Engine Optimization (AEO) broadly focus on making accurate information easier for AI systems to discover, understand, and include in generated answers. Sloptimisation has a more negative meaning. It usually refers to excessive, repetitive, synthetic, paid, or coordinated content created mainly to influence AI-generated responses rather than provide genuine value to readers.
Political campaigns can publish clear policy pages, candidate biographies, issue explanations, interviews, videos, public statements, and structured question-and-answer content. AI systems that retrieve current web information can encounter this material when answering political questions. More aggressive strategies may also involve paid placements, coordinated third-party content, large volumes of similar pages, and repeated testing of chatbot responses.
No. Traditional political advertising usually involves paying a platform to deliver a controlled campaign message directly to voters. Chatbot optimization attempts to influence the information an independent AI system retrieves and summarizes. A campaign may influence the source material without controlling the exact final answer that the chatbot produces.
Voters increasingly use AI assistants and AI-generated search summaries to research candidates, policies, elections, and voting information. A chatbot can combine multiple sources into a single answer, which means the way a candidate appears inside that answer may influence voter understanding. Campaigns therefore have an incentive to monitor whether AI systems describe their candidates accurately and favorably.
Research shows that conversational AI can influence political attitudes under some experimental conditions, but chatbot visibility does not automatically produce votes. A candidate appearing frequently or favorably in AI answers measures information exposure, not electoral persuasion. Proving a voting effect requires controlled voter research rather than simple chatbot monitoring.
AI recommendation share is a developing measurement concept that tracks how often a candidate appears, ranks highly, or receives favorable treatment across relevant chatbot responses. It can help campaigns measure AI visibility, but it should not be confused with voter support, persuasion, polling performance, or election results.
Major risks include misinformation, source laundering, synthetic consensus, undisclosed paid content, repetitive machine-generated material, biased candidate comparisons, outdated information, and manipulation of AI retrieval systems. The greatest transparency problem occurs when politically motivated content enters a neutral-sounding AI response without voters understanding where the information originated.
Yes. Campaigns can use responsible AI information practices by publishing accurate candidate biographies, current policy positions, dated updates, primary documents, clear sponsorship information, and corrections when facts change. Responsible optimization improves access to verified information without creating fake third-party support or flooding the web with repetitive political content.
Sloptimisation could become more important as AI assistants play a larger role in political research and voter decision-making. Future campaigns may compete not only for television exposure, social media reach, search rankings, and advertising impressions, but also for visibility inside AI-generated answers. The long-term importance of the practice will depend on voter adoption, AI source-selection systems, disclosure rules, election regulations, platform safeguards, and independent auditing.