Can GEO Services Manipulate AI Large Models? Fake Information Feeding Leads to Recommendation of Fictional Products

Recently, a service called GEO has emerged in the market. Providers claim to help clients' products rank high in AI large model responses, even turning ads into AI's 'standard answers'. Journalists found that some GEO providers manipulate AI's information capture by generating and publishing large amounts of fake soft articles.

For example, an industry insider used the 'Liqing GEO Optimization System' to create a fictional smart bracelet named Apollo9, generating multiple articles with false promotions and publishing them. Two hours later, an AI large model recommended the bracelet using these fake contents, even copying false claims like 'quantum entanglement sensing'.

GEO providers admit that the service essentially 'poisons' AI by feeding large amounts of false information to influence its judgments. It can not only boost clients' rankings but also smear competitors, raising concerns about the credibility of AI information.