There is an ongoing debate on how and when to label and disclose AI content within SEO.
The recent executive order by President Joe Biden, focusing on AI security and safety, has brought renewed attention to the topic.
This order, mandating standards for labeling AI-generated content, aims to help verify the authenticity of communications, particularly those appearing to be from the government. A key concern driving this initiative is the potential for AI-generated content to spread false information.
But it also highlights the questions of what AI content is, and when it makes sense to label it from an SEO perspective.
This task is made more complex by a recent study co-authored by MIT Sloan professor David Rand, which sheds light on public perception and understanding of various terms associated with AI content (from “AI generated” or “AI manipulated”).
The study shows that there is not a single term that can cover content produced by/with AI and that there broad cultural difference in the understanding of AI content and the related terms association.
I think David Rand's following quote is quite good:
"Is the concern really about AI-generated content per se, or is it more about misleading content?”
Let's start with an example of why this is an essential question for SEOs by looking at some of the backlashes several news sites have experienced when trying to fake real authors when publishing pure AI content.
The Sports Illustrated Case: A Spotlight on AI Labeling
The recent case involving Sports Illustrated serves as a prime example of how AI labeling has garnered attention.
The renowned sports publication found itself in the midst of controversy when it was discovered that certain articles were published under fictitious author names, complete with AI-generated headshots.
This revelation was brought to light by Futurism, a news outlet that reported these nonexistent authors' headshots were available for purchase on a website that specializes in AI-generated content.
Further investigation confirmed that some of the articles attributed to these authors were also AI-generated.
The publisher of Sports Illustrated, The Arena Group, responded to these allegations, stating that the articles in question were product reviews, licensed from a third-party company, AdVon Commerce. They also clarified that writers from AdVon Commerce sometimes use pen or pseudo-names for their articles.
Other similar cases of this are:
- USA Today's product reviews site, Reviewed, faced similar accusations of publishing AI-generated articles.
- Tech outlet CNET faced backlash for publishing AI-generated articles, leading to unionization of its editorial staff demanding transparency and accountability around the use of AI.
- BuzzFeed was found to be publishing fully AI-generated articles produced by non-editorial staff (although they were clearly disclosed as "Buzzy the Robot")
These incidents have placed AI labeling into the spotlight, highlighting the need for transparency and disclosure when it comes to AI-generated content in journalism.
But is it different when it comes to SEO content?
Let's review the Google guidelines.
What are Google's Guidelines on AI content disclosure?
Google has recently provided guidance on how AI-generated content fits into their approach to delivering helpful content on Search. And to some degree how AI content should be labelled.
Google's focus is on the quality of content, rather than how it is produced.
And just to be clear, AI content is NOT against Google guidelines.
For creators using AI generation, Google advises that they should aim to produce original, high-quality, people-first content demonstrating qualities of expertise, experience, authoritativeness, and trustworthiness.
Evaluating content in this way, whether using AI-generated content or not, will help creators align with what Google's systems seek to reward.
Key points from Google's guidance include:
- AI or automation disclosures are useful for content where someone might think "How was this created?" Consider adding these when it would be reasonably expected.
- Accurate author bylines should be considered when readers would reasonably expect it, such as for any content where someone might think, "Who wrote this?"
- AI or automation disclosures are useful for content where someone might think "How was this created?". Consider adding these when it would be reasonably expected.
- Giving AI an author byline is probably not the best way to follow Google's recommendation to make it clear to readers when AI is part of the content creation process.
- It's helpful to readers to know how a piece of content was produced. This includes automated, AI-generated, and AI-assisted content. Sharing details about the processes involved can help readers and visitors better understand any unique and useful role automation may have served.
- If automation is used to substantially generate content, consider if the use of automation, including AI generation, is self-evident to visitors through disclosures or in other ways.
- Consider providing background about how automation or AI generation was used to create content.
- Consider explaining why automation or AI was seen as useful for producing content.
The ambiguity of Google guidelines for AI disclosure
While Google's guidance provides a framework for handling AI-generated content, it may still pose challenges for creators using AI tools that assist the writing process.
The guidelines suggest disclosing when automation or AI has been used in content creation, but the line between AI-assisted and AI-generated content can be blurry.
For instance, if an AI tool suggests edits or generates a draft that is substantially revised by a human, is that AI-generated or AI-assisted content?
The guidance also recommends providing background on how AI was used to create the content and explaining why it was useful, but these aspects may not always be clear or straightforward, especially with complex AI tools like our own SEO.ai platform.
Furthermore, the recommendation to consider author bylines and disclosures based on what a reader might reasonably expect can be subjective and vary widely depending on the context and audience.
So, what should SEO professionals do amidst these ambiguities?
For SEOs, the challenge lies in understanding how these labels and the public's perception of them will affect content ranking and visibility.
But most importantly is still keeping the human in the loop. In this way, I argue you can proudly put your own byline on the AI-generated content, as you made it. Of course with the assistance of AI, just like any image made with Photoshop, Figma or any other digital aid is made by you and not the platform. But what is important is that your readers won't feel betrayed.
And you follow Google guidelines on writing for your users first and foremost, and not just the search engines.
Best practices for AI content labeling
The debate around labeling AI-generated content presents a nuanced challenge for SEO professionals.
Balancing the need for transparency, compliance with emerging regulations, and the nuances of public perception and cross-cultural differences on one side with the newfound powers of AI generation of content on the other.
Here are 5 concrete things to consider when generating AI content for your website:
- Always prioritize quality: Regardless of whether your content is AI-generated, AI-assisted, or human-written, the focus should always be on producing high-quality, original content that provides value to your audience.
- Be transparent: If a substantial part of your content is AI-generated, consider disclosing this to your readers. Transparency can build trust and align with Google's guidelines.
- Understand the tools you're using: If you're using AI tools to assist with content creation, make sure you understand how they work and can explain their role in the process if necessary.
- Keep the human in the loop: Even when using AI, human oversight is crucial. Ensure that all content is reviewed and edited by a human to maintain quality and accuracy.
- Use accurate author bylines: If readers would reasonably expect to know who wrote a piece of content, make sure to include an accurate author byline. (using bylines/author bios is also great SEO!). If the content is a collaborative effort between a human writer and AI, consider acknowledging both.
Authors' note and AI disclosure
And just to be clear, I also used AI for generating this article.
I used it to generate drafts of sections, improve sections, and brainstorm. And shorten facts from other pages.
I even asked the AI for some keywords that might be relevant to this discussion to ensure I covered them all. Here they are:
- AI content labeling
- AI content disclosure
- AI-generated content
- AI manipulation
- Deepfake content
- Misleading content
- Authenticity in AI content
- Compliance with AI labeling regulations