ChatGPT and OpenAI are making AI generated content accessible to anyone. But should marketers and reputation managers use it to create content? And if they do, what does it mean for the integrity of Google search results?
It’s well known that Google is by far the most used Internet search engine out there – as of March 2023, 92.63% of Internet searches are on Google. Bing comes in a very distant second at under 3% of Internet searches.
As reputation management experts, we also know how important the first page of Google search results is to the user. Now that AI generated content is becoming more mainstream, it’s vital that we understand Google’s stance on this in order to assess how it will impact search results.
Google’s take on AI-generated content using ChatGPT
Google’s public take on AI-generated content has changed over recent years. We still don’t know exactly how Google detects whether content is AI content as opposed to that created by humans.
There’s no doubt that it’s possible to identify whether content is AI created. For example, Google could develop and implement a system to detect this kind of content. However, it’s not clear whether this is happening right now, and there is no official evidence that it is.
Back in April 2022, Google senior search analyst John Mueller said in an interview that he can’t claim that Google could differentiate between content created by ChatGPT and human beings. This kicked off a long running debate over the comparison of AI content and spam, which remains outlawed by Google.
Fast forward to May 2023 and Google is far more open about accepting AI created content and what it considers to be spam or high quality.
The dilemma of AI created content for Google
Precisely because Google has always been about presenting users with the highest quality search results, the addition of ChatGPT content presents a new challenge.
Google has always been clear about its mission to find new ways to prevent spam from flooding its search results. And it has already been fighting the scourge of automated content clogging up search results for many years.
The platform defines auto-generated spam as: “… content that’s been generated programmatically without producing anything original or adding sufficient value; instead, it’s been generated for the primary purpose of manipulating search rankings and not helping users.”
Google has always been clear about its mission to eliminate this kind of automatically generated spam due to its low quality. Content like this usually consists of keyword stuffing designed to manipulate Google search results and, as such, the platform wants to remove it from SERPs (search engine results pages) in order to maintain the integrity of its results.
AI created content versus human created content
In the past, automatically generated content has been low quality. At the same time, Google has always been clear that human generated content must be high quality. In other words, all content – regardless of its origins – must be written in order to improve user experience and present them with high quality information.
More recently, there has been an increase in the use of machine learning (ML) and GAN (generative adversarial networks). The latter is a model that uses deep learning to boost the validity of the predictions of two neural networks in order to create targeted, high quality text.
Now that this kind of AI software is available to marketers and content creators (in the form of ChatGPT and Open AI), the line has blurred for Google’s quality guidelines. This is because it is now possible to use automated systems to create high quality content that is considered useful for the reader/searcher.
What does Google say about AI generated content?
In February 2023, Google’s Public Liaison for Search Danny Sullivan explained the company’s position on AI content, which is that Google accepts it as long as it is useful for the user and created for the user.
Therefore, if ChatGPT is used to create spammy content, it will violate Google’s guidelines and policies. If it used to create useful, high quality content, it will not. This brings it in line with the way that Google analyses and ranks human generated content too.
Google has updated its guidelines to cover both AI and human generated content, which includes the following suggestions:
- Who created the content – Google says it should be obvious through a byline or other explanation.
- How the content was created – Google says how the content was created should be obvious, whether it’s AI generated or AI assisted or written by a person.
- Why the content was created –Google specifies that the content should only be created in the first place if it’s useful to people searching.
Google is not anti-AI content
Officially Google has not yet explained how – or even whether – it’s detecting AI and automated content. However, it is clear AI generated content is not against the platform’s guidelines. As long as the ChatGPT content hasn’t been created in order to manipulate search rankings, then Google is basically OK with it
From Google’s point of view, the threat of poor quality or spammy automated content is not a new challenge. Rather it forms part of the platform’s continuous efforts to ensure the integrity of the search results.
In a similar way, Google says that it will continue to adapt its existing procedures and processes to analyse content for fake news as well as low quality. The systems continue to evolve and improve.
Google’s systems are the same for all content, whether it’s written by ChatGPT or by human beings. Google goes on to says: “Our focus on the quality of the content, rather than how content is produced, is a useful guide that has helped us deliver reliable, high quality results to users for years.”
How will AI content impact SEO?
It’s not fully clear how this will impact SEO and content creators – the message for now seems to be as long as ChatGPT is used to create high quality, targeted, useful content, it should be fine by Google. Whether rankings will be massively impacted remains to be seen.
The likelihood is that the next few years will clarify exactly how much ChatGPT and AI will change the way we use Google.
ChatGPT and online reputation management?
Online reputation could be impacted by generative AI, such as ChatGPT in a couple of key ways. For example, AI is changing the way in which content is displayed on search engines. Furthermore, the use of AI programs to create or enhance content, profiles and blogs in order to curate an online image, could backfire.
AI will increasingly integrate into search engines and web browsers and both of these scenarios are possible. And both of them would be damaging to the online reputation of a brand, business or individual.
AI created content could be reputationally damaging
Even if a possible client has already had some form of introduction to a business, they will still Google the brand name. And this certainty is precisely why an online presence is important.
Online reputation management is about creating, protecting and maintaining a positive – but honest – online reputation. Brands that choose not to have an online presence at all are unwittingly damaging their reputation too.
And while AI can, of course, be used to either create or update content, it’s worth understanding that we are a long way off from generative AI being able to replicate he kind of nuanced analysis within writing that humans can carry out.
The dangers of using generative AI such as ChatGPT for content creation
As yet, we wouldn’t advise using ChatGPT or similar programs to write content that will form the basis of a company’s online reputation. Using it to create base articles that can be amended and updated can be worth doing, as there are obvious benefits in saving time and money.
But it is incredibly risky to use ChatGPT without a professional content creator or editor thoroughly reviewing the results before publication. As it stands, the information and answers that AI regurgitates can be incorrect. They could, for example, be scraped from old content that is incorrect or negative.
Use of this will then go on to create a blog or piece of content that perpetuates the negative bias of the source content. This will be reputationally damaging.
The bottom line is that ChatGPT and associated tech is still so new that it’s largely untested in crucial ways. Over time, it will change and it’s likely that this will also change the way we search online.
But it also has a long series of challenges if it’s used for reputational purposes, and these should be considered by users.
Managing Director at Igniyte – The Reputation Experts
Roz is an industry spokesperson on all areas of online reputation management and our resident digital media expert. She regularly writes about reputation management research, online reputation risk and industry best practice.