Expert: AI Model Poisoning is a New Form of Unfair Competition

BlockBeats News, March 15th. The "3·15" Gala reported on the chaos of AI large models being "poisoned." Li Fumin, an expert from the Shandong University of Finance and Economics Institute of Societal Governance Intelligence, stated that in response to businesses conducting targeted training on large models through GEO and other services, guiding AI to generate specific product or service recommendations is essentially a new form of unfair competition and consumer misleading behavior that uses technical means for covert marketing and fabricating facts. This behavior leads consumers to unknowingly receive implanted marketing content, and its harmfulness and illegality require high attention. On one hand, the above behavior violates consumers' rights protection laws, which stipulate consumers' right to information and fair trade. On the other hand, it involves using technical means for false or misleading commercial propaganda, disrupting the normal recommendation algorithm order and market competition environment, thus constituting unfair competition.

Addressing the governance of the aforementioned AI poisoning behavior requires a multi-pronged approach. Regulatory authorities should prioritize monitoring AI-induced marketing and strengthen law enforcement supervision. AI operators should enhance corpus source review and output filtering, establishing a traceability mechanism. Consumers should increase their awareness of the commercial nature of AI-generated information and actively protect their rights through complaints and reports. (China News Service)