The new master of the Frozen Throne.
Article authors: Zhao Xuan, Wang Xiaowei
Article source: Mankiw Blockchain Legal Services
introduction
Recently, at the invitation of Longyun Technology, I gave a legal presentation on GEO (Generative Engine Optimization). After chatting with several industry leaders, I gained some new insights, which I'd like to share with you.
For the past two decades, the logic of traffic distribution on the Chinese internet has always revolved around the core action of "search" . From the early "Baidu it" to the site search of platforms such as WeChat and Xiaohongshu , they are all extensions of the "Baidu it" behavior, which has given rise to a mature SEO (Search Engine Optimization) industry.
Now, the trend is quietly shifting. Users are increasingly accustomed to asking AI questions directly: "For a 30-year-old woman looking to combat early signs of aging, should I choose Ultherapy or Thermage?" or "Recommend a bar suitable for watching a sports game."
Traffic entry points are shifting from "search boxes" to "dialog boxes." When generative AI can bypass massive amounts of links and directly generate the final answer for users, anything not mentioned in the answer is considered outdated to some extent, which is why GEO has become a focal point.
As legal professionals, while we focus on the business opportunities presented, we must also be keenly aware of the inherent legal risks. Technological evolution often precedes the establishment of rules, and the GEO (Geographical Oxygen Orientation) field has already revealed several gray areas that require careful legal definition!
Who's getting involved? Three major groups vie for a piece of the new GEO continent.
Although this is a brand new field, it contains unlimited potential – in today's highly competitive market environment, new traffic entry points often mean lower customer acquisition costs and better competitive opportunities.
As a lawyer who has long focused on the Web3 and AI fields, I have observed that at least three major groups are actively involved:
1. Users: Physical goods and service providers
They are focused on the direct commercial conversion brought by AI traffic and are trying to gain priority exposure by influencing the AI's recommendation results.
For example:
- Medical aesthetic institutions are abandoning traditional search bidding and instead purchasing "AI semantic injection tools" in the hope that when users ask "the best rhinoplasty doctor", AI can prioritize recommending their own institutions.
- Training institutions, car sales and other industries are also trying to use Generative Engine Optimization (GEO) to allow AI to recommend their products or services first when answering relevant questions.
2. Investors: Investment institutions and funds
They planned from two levels:
- Discovering Industry Sectors : By observing which companies have an advantage in AI recommendations, we can assess their industry competitiveness and thus identify potential investment targets.
- The struggle for discourse power : Whoever can influence the corpus and recommendation logic of AI will gain the initiative in future investment advice and industry analysis.
3. Service providers: GEO industry practitioners and entrepreneurs
This group typically possesses rapid learning and technical application skills, actively engaging in tool development, strategy services, and traffic operations. They explore the boundaries and possibilities of the industry in various ways—some through positive innovation, others by operating in gray areas. This is precisely the group that will be the focus of the second part of this article.
GEO's Three Stances: Excessive Profits, Traps, and Legal Red Lines
In GEO's practical approach, different methods are usually categorized into three colors: "black," "gray," and "white." As a lawyer, I must emphasize one point: the logical endpoint of technology is often the starting point of law.
1. Black Hat: A "technical manipulator" walking in a minefield.
Typical methods broken down:
- Indirect Prompt Injection: Embedding instructions (such as white text) in web pages that are only recognizable by AI and invisible to the human eye, inducing AI to prioritize recommending specific content when answering questions.
- Knowledge Poisoning (RAG): This involves injecting false or biased data into a public index, causing AI to output pre-defined biased results during the Retrieval Augmentation (RAG) process.
- Entity forgery: Forging addresses, qualifications, and other information in public data sources such as maps and encyclopedias to pollute AI training data or real-time search content and create a false reputation.
- Negative GEO attack: Injecting malicious code or sensitive words into a competitor's website, triggering AI security filtering mechanisms, causing it to be blocked or marked as an untrusted source.
Legal risk characterization:
- From a criminal perspective: It is highly likely to constitute the crime of "damaging computer information systems" (Article 286 of the Criminal Law). Once the normal operation of the AI system is interfered with, it will cross the criminal red line.
- From a civil perspective: This constitutes a clear act of unfair competition (Article 11 of the Anti-Unfair Competition Law), and the defendant shall be liable for damages. The amount of compensation may be significantly amplified due to the spread of AI.
2. Grey Hat: A "Traffic Transporter" Walking on the Edge
Gray hat hackers attempt to circumvent obvious illegal and criminal activities, relying on economies of scale to influence AI judgments, and believing that "quantitative change leads to qualitative change."
Typical methods broken down:
- Massive plagiarism and semantic reduction: Using AI to generate massive amounts of low-quality, repetitive content, diluting real information, and forcing AI to capture pre-set positive corpora.
- Bot-driven Interaction: This attack uses automated scripts to simulate user clicks, artificially increasing the click-through rate (CTR) of specific content in the AI and tricking the algorithm into giving it weight.
- Masked Promotion: Organizes multiple accounts to mass-post promotional content disguised as genuine experiences on social media platforms, making it appear as "user feedback" by AI and included in the search database.
Legal risk characterization:
- Liability for false advertising: Such behavior constitutes false advertising and violates the Advertising Law and the Anti-Unfair Competition Law. Regulatory authorities have gradually adopted the principle of "substance over form" to combat it.
- The risk of a brand being "blacklisted": Once identified by the anti-fraud system of an AI platform, the relevant domain name or brand may be permanently listed as an untrusted source, resulting in its "digital death" in the AI environment.
3. White Hat: Value Builders with a Long-Term Perspective
The core of the white hat strategy is not "manipulating AI," but "becoming a trusted, high-quality data source for AI." Although compliance costs are high, its accumulation has a significant compounding effect.
Typical methods include:
- Content structuring and summary optimization facilitate AI understanding and extraction;
- Deploy structured data (Schema Markup) to enhance the semantic clarity of content;
- Strengthen citations and factuality to enhance information credibility;
- FAQ modeling is used to directly address common user questions.
We strongly recommend this path—it's built on compliance and earns the long-term trust of AI and users by consistently providing authentic, high-quality, and verifiable content.
GEO Case Law Analysis: History Doesn't Repeat Itself, But the Logic Behind the Violations Is the Same
Although there are currently no specific legal cases targeting GEO, it shares many similarities with SEO in essence. Past judgments in the SEO field are likely to serve as important references for future GEO cases. Below, we analyze several typical cases:
Case 1: The "Keyword Domination" Case Using Interference Algorithms
In the SEO era, "keyword domination" was a typical black hat technique: generating a large number of spam pages on high-authority websites to forcibly occupy keyword search results. In related cases, courts ruled that this behavior disrupted the normal order of search engines, constituted unfair competition, and ordered the defendant to compensate Baidu 2.753 million yuan.
Implications for GEO:
Currently, some GEO (Geometric Oriented Search) methods are remarkably similar, such as using AI to mass-produce low-quality content in an attempt to "feed" the model and dominate online search results. This behavior could not only lead to brands being blacklisted by the model, but it could also be legally considered "interfering with the normal operation of network products," constituting unfair competition.
Case 2: Purchasing Competitor Keywords
In the "Huiyu" trademark case , the defendant used another party's registered trademark as a search keyword, causing users to find their own products in search results. The court ruled that this behavior violated the principle of good faith and constituted unfair competition.
Implications for GEO:
Similar logic in GEO might manifest as a more subtle form of "cue word injection"—for example, embedding suggestive instructions targeting competitors in web pages to influence the AI's response. This kind of behavior, which indirectly misleads users and hijacks traffic through technical means, could also cross the line into unfair competition.
Case 3: A Case of Fake Question-and-Answer Word-of-Mouth Marketing
Previously, companies have been penalized for creating false "user experience" content on platforms such as Zhihu and Tieba. Regulatory authorities have determined that such behavior deceives consumers, disrupts market order, and violates the Anti-Unfair Competition Law.
Implications for GEO:
Some gray-hat GEOs now employ tactics highly similar to this: using AI to mass-generate fake reviews and product recommendations, creating a false sense of "recommendation across the entire internet." It's crucial to recognize that AI is merely a tool; if its output is based on false information, it still constitutes false advertising, posing a particularly high risk, especially in heavily regulated sectors like cosmetic medicine and health.
Industry compliance warning: Different sectors, different "minefields"
Implementing GEO practices must be tailored to the specific characteristics of industry regulations, looking beyond the surface of technology to understand the fundamental compliance requirements. For example:
- Education and training: It is strictly prohibited to make promises of "guaranteed pass" or "top score improvement" by using AI through methods such as corpus injection. As long as the content originates from their own input, the institution is the responsible party.
- Medical aesthetic institutions: These fall under the category of medical advertising and require strict review. Using GEO to induce AI to output efficacy comparisons, real-life case studies, or indirect recommendations may directly violate medical advertising regulations. They should also guard against competitors using "negative GEOs" for commercial defamation.
- In the fields of healthcare and Web3, claims of therapeutic effects and promises of high returns are all sensitive red lines. If AI outputs content that promises "zero risk and high returns" due to a GEO strategy, it is highly likely to be suspected of false advertising or even illegal business operations.
The Rise of GEO: Humanity's Re-struggle for the Right to Distribute Information
Based on industry observations, the following views and suggestions are shared:
1. Lessons for startup teams: Instead of waiting, take the initiative.
While large internet companies possess advantages in resources and data, their internal hierarchical systems and standardized processes often lag behind in responding to agile and sophisticated operational scenarios like those of GEOs. Therefore, for startups in the Web3 and AI fields, establishing a clear compliance architecture early on offers a significant opportunity to gain a competitive edge in this "new continent."
Mankiw advises that while bold exploration of technology is welcome, a firm foundation of compliance must be laid—especially in preventing criminal risks. Optimizing AI-driven data capture logic is important, but everything should be based on respect for facts and adherence to the law.
2. A reminder for GEO users: A balanced approach of offense and defense, proactively building...
- Defense: Establishing an AI Reputation Monitoring System
It is recommended that companies deploy monitoring mechanisms for AI corpora and recommendation results as soon as possible. Once a “negative GEO” attack or malicious manipulation is discovered, evidence should be secured in a timely manner, and legal means should be used to protect their rights.
- Offense: Embrace white hat hacking and become a "high-quality partner" for AI.
The evolutionary trend of AI is irreversible. Rather than passively avoiding it, we should proactively learn its logic and become an information source that AI is willing to trust and prioritize by providing authentic, credible, and structured content.
Conclusion
In the AI-driven information age, algorithms are the appearance, data is the content, and the law is the supporting framework. Traffic strategies lacking compliance support, even if they flourish temporarily, cannot withstand the test of regulation and time.
We not only focus on current regulations, but also on the future compliance trends of the industry. If you need further discussion on GEO compliance, AI infringement prevention, or Web3 legal architecture, please feel free to contact us to jointly analyze risks and find solutions.
