In mid-May, Google unveiled a fresh set of governance guidelines tailored for generative search, categorizing malicious manipulation within Generative Engine Optimization (GEO) as 'spam content' to tackle the rising concern of 'AI poisoning.' The term 'AI poisoning' denotes the practice where advertisers craft deceptive marketing materials and disseminate them via SEO tactics, misleading large models into indexing false information as factual data, thus jeopardizing the integrity of Google Search. To counteract this menace, Google has rolled out three punitive strategies: immediate demotion in search rankings, AI-driven content removal, and comprehensive site bans. Nevertheless, industry experts express skepticism about the complete eradication of this issue, citing hurdles such as the clandestine nature of sophisticated content injection, the routine nature of attack-defense cycles, and the indistinct line between semantic interpretation and factual accuracy. Looking ahead, the emphasis in SEO will pivot towards substantiating content authenticity and factual sourcing, while search engines will be tasked with maintaining a delicate equilibrium between model adaptability and the purity of information streams over the long haul.