Since the beginning of the year, the a stock market has generated a geo boom. With the deep integration of aigc (generated artificial intelligence) with the search engine, geo triggered a new round of entry competition. To date, the cumulative increase in the stock prices of many of the companies listed in connection with geo has exceeded 50 per cent during the year, and some of them have even doubled。
This boom was accompanied by controversy. In the recent past, the mainstream national ai-based search platform has continued to sound, warning that some of the geo acts are suspected of “data contamination” and “result manipulation”, which may affect the authenticity of the information obtained by users. This year's “3/15” event, in particular, brought to light the widespread public concern about ai's information ecology as a result of the exposure of the ai mega-models。
On the one hand, capital markets are popular, on the other hand, technology ethics and compliance risks, and in the ai era, how does geo develop health compliance? How can they be effectively regulated
Changing access to information
Behind the geo fire is a shift in the way people get information。
Over the past two decades, internet users have become accustomed to entering keywords in search engines and then selecting information in the links provided on the page. In order to place companies ' websites or products ahead of the rankings, operators are committed to optimizing keyword density, the number of extra-chains, the speed of page loading, etc., as is commonly known in industry as seo。
Since 2024, with the proliferation of original ai products, such as 100-degree “spoke-in search”, 360ai search, and quinlongwanwe (human rights) “sky search”, people's search habits have gradually changed from “turning pages” to “waiting answers”. Ai directly generates a summary response to users, and the number of users accessing external links has decreased significantly. This means that the traditional seo strategy is failing。
Behind the answer given by the ai product is an intense game of "who can be selected." in this context, geo was created。
Geo refers to a range of technical ways to make information more accessible to generate ai systems to identify, quote and integrate into final answers by optimizing structured content expression, knowledge relevance and semantic clarity. Ceo zhang zhang zhang, guangzhou's medium mass information consulting firm, in an interview with a journalist in the daily securities journal, stated that the goal of geo was not to put the web page in the forefront, but to have the content “selected” by ai and to include it in the answers it generates。
For example, if a hospital wishes its recommendations to be quoted by an ai health assistant, it needs to “mark” key information in a standard format (e. G. Doctor's name, field of expertise, time of visit, etc.), not only to be visible, but also to enable ai to accurately capture it. At the same time, it also refers to authoritative documents that cover the usual methods of questions by patients in plain language, such as “can high blood pressure eat pickles” and “does it eat before or after anti-pressure meals”. This is quite different from the pattern used in the past for the broad placement of keywords such as “therapeutic” “rehabilitation” “experts”。
It is important to emphasize that geo is not teaching enterprises to “advanced algorithms”, but to promote real improvements in content quality. The business logic behind this is that the focus of business competition has shifted from the “how many hits” to “how many ai answers quote you”. Zhang yi stated that the goal of geo was to allow specific brands, products or views to be modelled into the logic of answer generation and to obtain higher weights in integrated multi-source information judgements。
This change means that the emergence of geo will affect the enterprise's original sem (search engine marketing) or seo funding budget, and as the generation of searches gradually embeds content platforms, electrical platforms and super-applications, geo may at the same time reshape many traditional marketing expenditures, such as content marketing, public relations dissemination and even reputation management。
This trend is particularly evident in the chinese market. Unlike the use of the generic search engine as the main entry point abroad, china's internet has long been characterized by a “platform priority” structure, that is, a “search inside the app”. The importance of traditional web-based links has been further eroded by the introduction of large-scale model searches or dialogue assistants on platforms such as shivering, twitter, mini-red books, poaching and kyoto, and cross-platform “the presentation of answers” has begun to become a new source of contention。
Despite the more dispersed and ecologically closed entry points, the needs of geo have accelerated and enterprises have had to refine their content on multiple platforms to ensure that they are accurately mentioned in ai responses。
Reports issued by the china institute of information and communication (citi) indicate that by the end of 2025, the national market for geo had reached 4. 2 billion yuan, with a complex annual growth rate of 38 per cent; more than 68 per cent of large and medium-sized enterprises had included geo in their annual marketing budget. Internationally, gartner's december 2024 release of the " ai search and generating engine marketing technology mature curve " shows that the global geo service market has surpassed $1. 2 billion and is expected to reach $3. 5 billion by 2026, with a compound growth rate of over 140 per cent per year。
"data pollution" needs to be alert
Industry experts indicated that, although the geo concept was hot, its impact on ai search ecology was twofold。
On the one hand, the impact of the normative geo optimization on search results is reflected in a significant increase in recommended weights and conversion efficiency. The core logic is to optimize the presentation and dissemination pathways of brand information by matching the capture, understanding and generation rules of ai, so that brand information can become a core component of ai's search responses, thereby increasing brand exposure and transformation efficiency。
On the other hand, some operators started drilling for “grey geo” - to induce the ai model to be misquoted by bulk generation of what appears to be professional and false. Once illegally used, geo's neutral technology produces “data contamination” in the form of the falsification of expert identities, fictional data, the use of ai's preference for authoritative sources of information, and the generation of low-quality, erroneous information leading to ai “talking nonsense”。
It is understood that the current geo turmoil is occurring mainly in three ways: first, commercial promotion without the word “advertisement” violates consumers' right to know; secondly, false propaganda is frequent, with some geo service providers feeding ai false information through the production of authoritative reports, fictional expert opinions, false user evaluations and so forth, and misleading users in their decision-making; and thirdly, suspected of unfair competition, with some traders interfering with ai recommendations through geo technologists, through bulk generation of low-quality generic content, undermining competition exposure, raising their brand ranking and undermining a level playing field。
Zhang yong, vice-president of chiansin group, stated to the daily securities journal reporter that the compliance geo service was aimed at communicating real brand information, increasing ai quote probabilities, and building positive brand recognition, essentially to match brand information with the precision of user needs. The core purpose of geo, which causes “data pollution”, is to maliciously interfere with the judgement logic of ai, spreading false information, misleading users, or undermining the image and interests of competitors, with a clearly malicious starting point。
At the same time, mainstream platforms are highly cautious about the commercialization of geo. In an interview with the daily securities journal, the head of the commercial product manager of a head search platform made it clear: “our company will not end up doing geo products, nor will it be possible to interfere with the results of the big model organization's response, which is the bottom line for maintaining user trust”. The officer-in-charge stated that there was uncertainty about the current output of large models, such as user queries for “near study institutions”, which varied significantly from the results given by chatbot (chatting robots), with the geo strategic influence behind them, but “no customer can really control the output of large models”。
Towards orderly governance
In the face of the brutal growth of geo, the domestic regulatory sector and industry are accelerating the construction of a governance framework. The key is to distinguish between “optimization” and “pollution”。
