Numerous writers have pointed to the problem of AI-generated slop polluting the internet. As a cultural phenomenon, it not only threatens the to drown genuine creativity, but arguably even undermines AI itself.
Jason Weiland writing on Medium expressed his frustration with what’s become of the internet. “I love AI. I use it every day. But I’m exhausted. … Everything is AI now. And I can’t tell what’s real anymore.”
According to an Axios report, over half of all content on the internet is now AI-generated.
Distinguishing between AI-generated content and content created by humans is only going to get harder with the proliferation of tools to humanise AI content. Popular tools such as Walter Writes AI list Reuters and The Guardian among their clients.

This clearly is highly problematic for many of the belles artes from journalism to academia, where high-quality writing is a measure of the authoritativeness of the underlying work, whether it is investigating crime and analysing ancient manuscripts. And it is forcing editorial review boards to use AI tools to help distinguish between human authored and AI authored work.
What is lacking in this space is a means of self-certifying work in easy-to-understand categories, and it is with this in mind that I conceived several years ago of the Authorship Classification Scheme, which I have now systematised into five categories. These categories are a measure of the contribution of AI to the creation of a particular piece of content, and I hope that they will prove sufficiently flexible to be used with the widest range of content, including textual, audio, video, visual and physical forms thereof.
This website will provide easy-to-use identifiers to allow anyone to classify content by its authorship.
If you are providing a simple prompt to Gemini or ChapGPT, then tag the content AI Auth+. If generative AI is creating content which appears in the final production, even if in part, then this is AI Auth. On the other hand, if you are only using AI for background and research, and not for, say, the writing of a final text, and no AI-generated text appears in the final text, then this would be AI Prompt.
Now to the present article. It’s taken me about 30 minutes to write, and despite AI Overview popping up when I do a Google search for sources, I have avoided it, and stuck to internet search results. Although these seem to be written by human authors, as most are post-2022 (given the nature of the topic), I really have no way of knowing whether they are AI slop or the author’s own thought and research, so I would have to classify this article as Human Auth.
The gold standard for human authorship, Human Auth+, is when you have not only researched and written the piece, or created the content, without any assistance from AI, but the work is either entirely the product of your own imagination, or you have used only non-AI sources to write it. This could either be text written prior to 2022, or alternatively, other work tagged as Human Auth+, or the equivalent.
Please let us know if you find the classification scheme helpful, or would like to incorporate it into your own tools and products. We look forward to hearing from you.
David McHutchon