Is AI Content Bad For SEO? | The SEO Consultant.ai LLC

A Common Sense Approach to SEO’s Taboo Subject

With the rise of ChatGPT and other LLMs such as Gemini and Perplexity it’s of no surprise that one of the most debated topics in SEO today is whether AI-generated or AI-assisted copy is bad for SEO. For simplicity’s sake, I’ll mostly refer to ChatGPT here since that is the most widely-known and used AGI app amongst the general public today. So let’s dig in and see how we got here.

ChatGPT Timeline

November 30, 2022 – OpenAI released an early demo of ChatGPT (ver. 3.5) which quickly went viral on social media. (source: Forbes)
February 1, 2023 – OpenAI announced ChatGPT Plus, a premium subscription option for ChatGPT users offering less downtime and access to new features. (source: Search Engine Journal)
February 7, 2023 – Microsoft announced ChatGPT-powered features were coming to Bing. (source: Search Engine Journal)
March 14, 2023 – OpenAI releases GPT-4 in ChatGPT and Bing, which promises better reliability, creativity, and problem-solving skills. (source: Search Engine Journal)
March 21, 2023 – Google launched Bard, its ChatGPT alternative. (source: Search Engine Journal)
May 23, 2023 – Microsoft announced that Bing would power ChatGPT web browsing. (source: Search Engine Journal)
July 25, 2024 – OpenAI launched SearchGPT, an AI-powered search prototype designed to answer user queries with direct answers. (source: Search Engine Journal)

Since ChatGPT is based on LLMs (Large Language Models) it’s suited for generated text-based responses that for the most part can sound reasonably convincing and grammatically-correct. While I’m no engineer nor an expert AI researcher, it’s well-documented that GPT was trained in large part by crawling the web up to a certain point which as we all know includes both factual information and incorrect information. The internet is still very much like the Wild West in that anyone can publish anything so learning models like this can be tricked into believing false information to be true.

Common Uses For ChatGPT

With the increase in popularity for AI apps such as ChatGPT, everyone from students to working professionals like you and I use it to varying degrees. It’s quite easy to generate homework assignment responses and even write entire books with enough concise prompting. Some engineers event consult with LLMs to create python scripts or review code if they’re stuck on something or for increased efficiency. What we can do with this technology is pretty incredible already and it’s still in its’ infancy.

With that said, you can’t just take an output directly from ChatGPT and use it verbatim then expect it to be perfect for you. If you’re going to use or reference AI-generated content in your day-to-day life then I would highly recommend fact-checking and having some base level of knowledge of the subject before using it. For many subjects, you might get it 80% of the way there if you have good prompts. That 20% though can really hurt you.

AI-Writing Tools For SEO

A quick Google search reveals a number of AI-writing tools out there on the market for SEO. I recently joined the SMX Next 2024 event in November and it seemed like even Bruce Clay was promoting his new tool called, PreWriter.ai. The ones I have personally tested include:

Using AI Tools for Writing

Whether or not most SEO’s are currently using AI-generated copy is hard to gauge as SEO is pretty insular in that public figures will often say one thing then do another in private. I know this because I’ve attended a number of conferences over the years and have had private talks with well-connected people. It’s a competitive channel so it’s only natural that you probably wouldn’t share every single thing that you do for public consumption if you actually care about helping your companies or clients. But I will say that I have personally have used it for some projects so I’m just sharing my own views here and personal experience.

Rarely does a day go by when I don’t see someone on LinkedIn (I would just hide the feed if it was possible) or talk to someone who says that using AI is bad for SEO and will kill your traffic. My response to that is I think that point of view is myopic and lacks nuance. One SEO told me that if you use AI copy then Google will automatically detect it then penalize you. Other people have suggested that AI tools embed some secret code that makes the content automatically detectable. Dear Conspiracy Theorist: Please tell me if Lee Harvey Oswald acted alone. You can’t judge content as some monolithic concept.

I take a lot of things that Google says with a grain of salt but Google themselves updated their search quality guidelines on February 8, 2023 to specifically address this question. You can it here straight from the horse’s mouth on Google Search Central: Google Search’s guidance about AI-generated content.

Rewarding high-quality content, however it is produced

Google’s ranking systems aim to reward original, high-quality content that demonstrates qualities of what we call E-E-A-T: expertise, experience, authoritativeness, and trustworthiness. We share more about this in our How Search Works site.

Our focus on the quality of content, rather than how content is produced, is a useful guide that has helped us deliver reliable, high quality results to users for years.

For example, about 10 years ago, there were understandable concerns about a rise in mass-produced yet human-generated content. No one would have thought it reasonable for us to declare a ban on all human-generated content in response. Instead, it made more sense to improve our systems to reward quality content, as we did.

Focusing on rewarding quality content has been core to Google since we began. It continues today, including through our ranking systems designed to surface reliable information and our helpful content system. The helpful content system was introduced last year to better ensure those searching get content created primarily for people, rather than for search ranking purposes.

So let’s play back this point of view that Google will penalize a site for using AI-crafted copy. If I personally write a paragraph myself about a topic then go to ChatGPT or another AI tool to prompt for the same thing, then get 90% of the same copy back, does it really make sense that Google is going to penalize me for using some variation of the AI content? There should be little distinguishable difference for the reader if I’m doing this the right way.

One of the main limitations to LLMs currently is that the writing often lacks personality and there are some pet phrases that ChatGPT loves to use. This is why it’s important for a human to edit and review copy before publishing. That has always been true. The truth is that most SEO agencies have always outsourced some or most of their content writing to offshore sources or farmed it out to Fiverr or Upwork. But what you get back is rarely something you can publish verbatim. So that’s where a good editor comes into play and can craft that first draft of copy into something good. That’s no different from how I would suggest using AI-generated copy.

Where I think people get into trouble is when they start cranking out hundreds or thousands of pages in a short amount of time that have clearly been AI-generated riddled with factual errors and published without professional oversight. You can also have the same problem with programmatic “product-led SEO” generating countless pages of poor-quality content even without the use of AI. In the end, these are just tools to help but you still own the final output.

AI Content Templates For SEO

I’ve listed a number of different AI-based SEO tools above that can help you with crafting content. Most if not all of these start with creating a template that helps writers to get started with writing the content. There are some overlap between these tools while some have more innovative ways of going about doing this.

One of the more interesting tools that I have tested is SERPrecon, a new tool created by Ryan Jones (SVP of Razorfish), which uses vector embeddings and cosine similarity metrics from top ranking sites to create content outlines based around your target keywords. Another tool that stands out is inLinks which was created by Dixon Jones (ex-CMO of Majestic SEO). inLinks takes a different approach in that the tool seeks to create a knowledge graph from your site content then aligns that with entities that Google can recognize. Their AI writing tool uses a natural language processing (NLP) algorithm to create content outlines with headings based on your target keywords. If you’re inclined you can also have the tool auto-generate a first draft of content based on the outline, which is powered by ChatGPT’s API.

Ways to Leverage AI for SEO

In some ways, AI has been used for web content for at least several years now. That didn’t begin with ChatGPT. Language translation tools have been around for a while now. Some commonly used tools include Google Translate and DeepL. These are efficiency tools. Whether or not you feel these tools are effective or not is another story but translators are not typically thought of as spam.

You can also leverage Screaming Frog’s new integrations with OpenAI API and other LLMs to create product descriptions at scale. Just be sure to review before publishing! I’ve spent several hundred dollars of my own money testing Screaming Frog’s OpenAI API integration to auto-categorize thousands of item pages based on the site taxonomy. The results were nearly 100% accurate. The pages that were not accurately categorized was due to poor-quality or thin content so the topic was not clear. While categorization might not exactly be considered “content” I think it’s relevant to what I’m saying here.

No one is suggesting that you need to use AI tools like the ones I’ve described in this article to craft content for your website but there is a place for them as well. I’m clearly more open-minded on this topic than some people but I just see these as tools that can help with efficiency and research. It’s still up to us to decide what is or is not appropriate to publish on websites. Nothing has changed in that regard. In the end it goes back to what Google says. It doesn’t matter how you go about creating the content as long as it’s of high-quality and helpful for your users. So my point of view on “AI content” is that it’s an irrelevant question. Focus on what your target audience wants. Do whatever it is you need to do to accomplish that whether it’s by hiring more writers, using AI writing assistants, using a programmatic approach or some combination of all of this.