GEO Accused of "Poisoning" the Well, Brokerages and Funds Collectively Puzzled: Can We Still Do It? How Are Boundaries Determined?

What is the background of the AI · GEO poisoning incident that was exposed?

Financial Associated Press, March 16 (Reporter Wang Yuling) — “We are considering whether to invest in GEO. After the March 15 exposure, can we still do it?” a securities firm asset management professional told reporters. This morning, the company has already communicated with several vendors about the legality and compliance of GEO.

GEO, or Generative Engine Optimization, allows AI to quickly extract key information from content and prioritize referencing brand content when generating answers. According to reports, many securities firms and fund companies are experimenting with GEO technology to increase their brand exposure and mention frequency in AI-generated content.

After the “poisoning-type GEO” was exposed on the March 15 evening show, it sparked widespread industry discussion. Is normal GEO placement compliant? How should the boundaries between compliance and violation be defined?

Compliance optimization or “poisoning-type GEO”? Where is the line?

“I personally think that normal optimization, as long as the content is not illegal, is similar to content marketing,” said a senior manager of a large securities firm’s online finance business.

“If it’s normal GEO placement, it should be similar to the traditional search era, where brands compete for rankings on search results pages through SEO optimization, which is a marketing method,” said the aforementioned asset management professional.

So, how do we distinguish between normal optimization and “poisoning”?

A founder of a GEO vendor told the Financial Associated Press that he believes the “poisoning-type GEO” exposed by CCTV involves: first, content generated in bulk by bots; second, the information is fictional; third, dissemination through manipulation of numerous accounts. He believes that false information is the key bottom line for judging compliance.

In his view, legitimate GEO optimization must at least meet four points: “content compliance, data compliance, channel compliance, and qualification compliance.”

“If illegal acts such as fabricating and spreading false securities information are carried out through AI training or feeding, it is prohibited by law. From a regulatory perspective, the requirements are higher, and the process of discovering illegal acts may be more complex. Regulators need to keep pace with the times,” said Xu Feng, lawyer at Shanghai Jiucheng Law Firm. He believes that regulators have already taken notice of the situation, and legislation and regulatory measures will follow quickly.

A legal PR professional believes that the current chaos in GEO mainly involves three aspects: first, commercial promotion without marking “advertisement” or “promotion,” infringing on consumers’ right to know; second, frequent false advertising, with some GEO service providers fabricating authoritative reports, creating fake expert opinions, and forging user reviews to feed false information to AI, misleading user decisions; third, suspected unfair competition, where some merchants manipulate AI recommendation results through GEO technology, generate大量低质量、同质化内容，压低竞争对手曝光率，提升自身品牌排名，破坏公平竞争环境。

The aforementioned GEO founder believes that the root cause of the chaos is that after Baidu’s search dominance weakened, many practitioners simply transplanted their original SEO experience into the GEO field, but in this new track, effective regulation and industry standards are lacking, leading to increasingly aggressive tactics.

He thinks that governing industry chaos depends on the normative guidance of large model vendors. “Currently, their main focus is on model training. Once they have the capacity to regulate information sources in the future, industry order will gradually establish. For example, models in fields like medicine and law do not allow external data crawling, so such chaos is less likely to occur.”

What legal consequences might “poisoning-type GEO” face?

If someone indeed operates a “poisoning-type GEO,” what penalties could involved parties face?

The legal PR professional said that the parties involved in operating “poisoning-type GEO,” including the client company personnel and PR firms, could be suspected of false advertising. According to Article 222 of the Criminal Law, advertisers, advertising operators, and publishers who violate national regulations by making false claims about goods or services with serious circumstances can be sentenced to up to two years of fixed-term imprisonment or criminal detention, and may also be fined.

How is “serious circumstances” judged?

According to Article 67 of the “Standards for the Filing and Prosecution of Criminal Cases by Public Security Organs” issued by the Supreme People’s Procuratorate and the Ministry of Public Security, violations include:

• Illegal gains exceeding 100,000 yuan;
• Making false claims about food or medicine with illegal gains over 30,000 yuan;
• Even if the above thresholds are not met, if within two years there have been more than two administrative penalties for false advertising, and false advertising is repeated;
• Causing serious harm or adverse social impact.

As a new phenomenon, GEO currently has no criminal cases. However, according to news published on the China Court website hosted by the Supreme People’s Court, the People’s Court of Linxiang City, Hunan Province, once sentenced two defendants for misleading consumers with fake positive reviews.

The presiding judge also issued a “reminder”: forging positive reviews to gain false market reputation is essentially covert false advertising, which not only infringes on consumers’ right to know and choose but also damages fair competition, squeezing out honest businesses. Do not buy or provide services just to boost false credibility.

During the two sessions, Minister of Justice He Rong stated at the “Minister’s Passage” during the Fourth Session of the 14th National People’s Congress that this year they will further strengthen legal and regulatory provisions in key, emerging, and foreign-related fields, accelerate legislative research on artificial intelligence and low-altitude economy, and better align rule of law with reform, development, and stability. Relevant legislation in the AI field is expected to become more complete and clear.