A Google podcast underscores the risk of factual errors in AI-generated content, emphasizing the need for human oversight.
Google warns that generative AI is prone to producing inaccuracies. AI models could spread outdated SEO advice, posing significant risks. Human review and fact-checking are crucial before publishing AI-generated content.
In a recent episode of Google’s “Search Off The Record” podcast, team members explored using Google’s new AI tool, Gemini, for creating SEO-related content.
Their experiment, however, raised concerns about the accuracy of AI-generated content, highlighting the need for thorough vetting.
The discussion featured Lizzi Sassman, Gary Illyes, and John Mueller, who took turns using Gemini to generate sample social media posts on technical SEO topics. As they reviewed Gemini’s results, Illyes pointed out a significant issue inherent to most AI tools:
“My primary concern with nearly all generative AI is its reliability—you must always fact-check whatever it produces. It worries me that we might be sharing this live and potentially disseminating information that isn’t even accurate.”
Outdated SEO Advice Exposed
Concerns arose from an AI-generated tweet suggesting using rel=”prev/next” for pagination—a technique Google no longer supports.
Gemini, Google’s AI tool, recommended the following tweet:
“Pagination causing duplicate content headaches? Use rel=prev and rel=next to guide Google through your content sequences. #technicalSEO, #GoogleSearch.”
Lizzi Sassman quickly pointed out that this advice was outdated, while John Mueller confirmed that Google no longer recognizes rel=prev and rel=next: “It’s gone. It’s gone. Well, you can still use it, but Google ignores it.”
Earlier in the podcast, Sassman warned that AI-generated content might contain inaccuracies due to outdated training data. She noted that:
“If there’s enough misinformation circulating or an older SEO strategy that has been heavily blogged about, it might surface during our exercise today.”
It didn’t take long for that concern to materialize, with Gemini suggesting obsolete SEO practices within minutes. This reinforces the need for thorough fact-checking and human oversight when using AI-generated content.
Human Oversight Still Critical
The Google Search Relations team, while acknowledging the potential of AI-generated content, emphasized the importance of human fact-checking during their discussion.
Illyes’ concerns echo the broader conversation about responsible AI adoption. Human oversight plays a crucial role in stopping the spread of misinformation. As generative AI usage grows, it’s essential to remember that its output should not be accepted at face value and requires validation by subject matter experts.
Why We Care
Although AI-powered tools offer valuable assistance in content creation and analysis, Google’s team has shown that a cautious approach is necessary.
Using generative AI without thorough review could publish outdated or misleading information, potentially damaging your SEO and reputation. If navigating these complexities feels overwhelming, our monthly SEO packages are designed to provide expert guidance and support. Let us help you ensure your content is accurate and reliable.
