Update history
What official documentation confirmed in 2026
Google named attempts to manipulate generative AI responses in its spam policies, Naver published its AI Briefing citation criteria on its official blog, and Google stated that structured data isn’t required for generative AI search. Only items published in official documentation are listed, not claims relayed by industry blogs.
- 01What goes into the update history
- 02What changes in practice
- 01
Keyword matching
Engines checked whether a page contained the same string as the query.
The era when repeating the same keyword worked
- 02
Semantic search
Different wording with the same search intent came to be treated as the same query.
Matching keywords alone loses its effect
- 03
Entity recognition
Brands, places and concepts are grouped as single entities.
Inconsistent naming scatters entity signals
- 04
Generative answers
AI answers appear before the results list, and documents show up as cited sources.
Reading fewer clicks only as a drop in rankings is a misreading
Update history
What goes into the update history
Mixing claims relayed by industry blogs with official documentation would make the entire history unreliable, so only official sources make the list.
Google names manipulation of generative AI responses in its spam policies
Google’s spam policies document added attempts to manipulate generative AI responses in Google Search. Rather than a new rule, it means all of the existing policies apply to AI responses as well.
Source: Google Search spam policies (official documentation)
Naver publishes its AI Briefing citation criteria on its official blog
Naver wrote: “AI 브리핑은 기본적으로 기존 검색 결과에 잘 노출 되는 양질의 글을 AI 브리핑 요약에 사용할 가능성이 높습니다.” (Our translation: “AI Briefing is basically more likely to use high-quality posts that already appear well in existing search results for AI Briefing summaries.”) That’s why plans for Korea and for everywhere else need to be separate.
Source: Naver Search official blog (2026-05-26)
Google says structured data isn’t required for generative AI search
Google’s guide states: “Structured data isn't required for generative AI search, and there's no special schema.org markup you need to add.” It also says: “You don't need to create new machine readable files, AI text files, markup, or Markdown to appear in Google Search (including its generative AI capabilities), as Google Search itself doesn't use them.” Keep schema for rich results, and don’t count it as a citation tactic.
Source: Google Search Central, “Optimizing your website for generative AI features on Google Search” (updated 2026-07-10)
In practice
What changes in practice
Language to cut from proposals
Claims that assert AI visibility or citation counts
Google’s spam policies now state that they cover attempts to manipulate generative AI responses, and its generative AI guide lists seeking inauthentic mentions among the things that don’t help.
What you can cut from the budget
Schema or special-file work aimed at AI citations
Google has stated that structured data isn’t required for generative AI search. Keep schema for rich results, and treat llms.txt as a format for agents rather than search.
A new metric
Naver AI Briefing citation count
Currently about the only channel in Korea where citations can be confirmed as a number. We record the same metric at kickoff and at the end of the engagement.
This field turns over every six months or so. That’s why we list the last verified date alongside the official source, and state plainly when something couldn’t be verified. Proposals that reuse outdated numbers carry the most risk.
We build to the standards written here. First, we check where your current website is getting stuck.
Free audit