As you know, Google updated its Search Quality Raters guidelines with 11 new pages a couple of weeks ago. But with that, Google also added the term “generative AI” to the quality raters guidelines for the first time, and it is listed 14 times.
Now, while Google’s Search Quality Raters guidelines have mentioned AI in the past, it looks like Google is getting more serious about where AI is acceptable and not acceptable. The previous versions did not outright mention “generative AI” from what I can tell.
This was spotted by Adomas Šulcas who posted about this on LinkedIn (hat tip to Mike King for tagging me). Adomas wrote:
Google recently updated the Search Quality Rater Guidelines and included a few major changes, mostly in the domain of Generative AI.
There’s a total of 14 entries for “Generative AI” in the PDF (find the link in the comments). Outside of the singular definition, most mentions are negative.Most of them are warnings that generative AI is being used for scaled content abuse or for paraphrasing.
While they do mention a positive (“Generative AI can be a helpful tool for content creation, but like any tool, it can also be misused.”), it’s clear Google is being extra careful about any generated content.
If judged that generative AI was used for some form of misuse, the site quality in almost all cases is rated “Lowest”.
So, while not groundbreaking news, AI-generated content, especially at some reasonable scale, remains risky.
First, on page ten, Google defines generative AI:
Generative AI is a type of machine learning (ML) model that can take what it has learned from the examples it has been provided to create new content, such as text, images, music, and code. Learn more here. Different tools leverage these models to create generative AI content. Generative AI can be a helpful tool for content creation, but like any tool, it can also be misused.
Google then listed it as a method to deploy scaled content abuse:
Examples of scaled content abuse include: Using automated tools (generative AI or otherwise) as a low-effort way to produce many pages that add little-to-no value for website visitors as compared to other pages on the web on the same topic.
Pages and websites made up of content created at scale with no original content or added value for users, should be rated Lowest, no matter how they are created. Even if you are unsure of the method of creation, e.g. whether or not the page is created using generative AI tools, you should still use the Lowest rating when you strongly suspect scaled content abuse after looking at several pages on the website.
Here are more sections where Google uses the term:
Content that is paraphrased from a single source or multiple sources. Content from a single page or from many websites can be summarized, reworded or paraphrased by people or generative AI tools. Paraphrasing may be valuable, for example when an expert paraphrases the contents of a government policy in easy-to-understand language.
Important: Copying, paraphrasing, embedding or reposting content does not automatically make a page Lowest quality. However, the Lowest rating is required when all or almost all of the MC on the page (including text, images, audio, videos, etc) is copied, paraphrased, embedded, or reposted with little to no effort, little to no originality, and little to no added value for website visitors. Likewise, the use of Generative AI tools alone does not determine the level of effort or Page Quality rating. Generative AI tools may be used for high quality and low quality content creation. For example, a high level of effort may be involved in creating high quality original artwork using Generative AI tools. However, it’s also possible to use Generative AI tools to create Lowest quality content with little to no effort, little to no originality, and little to no added value for website visitors.
Paraphrased content can be much harder to recognize. Use the same process described above to search for other pages: Have words or other indications of summarizing or paraphrasing generative AI tools, such as words like “As an AI language model” (example)
Here are examples used with those terms:
Forum discussion at LinkedIn.