Do AI Detectors Flag Human-Written Content?

The goal of AI detectors is to spot content that's written by AI. The question is: do these detectors flag human-written content?

Turns out, yes.

And what does this mean for your SEO efforts?

Here are a handful of tests that I ran on content written by me, my fellow human writers, and Goldman Sachs.

Test 1: Commencement Address

First up, the commencement speech I gave at my alma mater, the University of Maryland, back in 2011. This predates generative AI. It took me about three weeks to write and revise. I got input from a dozen trusted colleagues before delivering it to the graduating class.

AI detectors said 12% was written by AI.

This falls into their "likely human written" result basket.

Test 2: Thought Leadership Articles and Posts

I tested articles that I wrote along with some that one of my writers drafted for a project. He’s an excellent writer and has done many assignments for me. I know he wrote the piece I because it flowed very closely with an interview I did with the client. All the articles I tested were 100% human written but the AI detector flagged parts of it.

These articles scored in the range of 15-40% as AI-written.

Checking this on a LinkedIn post I wrote, the AI detector had some mixed opinions (quick note: I write 100% of what you see me post on LinkedIn, no AI).

Test 3: Popular Goldman Sachs Report

I wanted to check this against industry reports from reputable organizations to see how well they fared. Here’s a look at a popular one, Goldman Sachs Creator Economy Report which is frequently cited for its prediction that the creator economy will be $500 billion by 2027.

This well-written report that includes data clocked in with an AI detector score of 27% AI-written. Big time false positive.

What This Means for Your SEO

The false positives from AI detectors are worth noting. Even the top organizations in the word have human-written content partially detected as AI generated. I tested a few other long articles as well and got similar false positives, which leads me to something I consider to be an important finding:

Long-form human-written content consistently has portions flagged as AI

AI detection algorithms almost always score long-form content as having some degree of AI-generated text, even when they are 100% written by humans. I presume this is because if it finds a match with a typical GPT phrase, it thinks AI wrote it.

Google uses its own detection algorithm. Although nobody knows its secret sauce, it's likely they know that false positives are part of the analysis. From what I can tell, they are deprecating the rankings of sites that contain content that is obviously written by AI, which means it's okay to have some degree of "AI written" in your AI detection score.

Based on these non-scientific tests, I would view an AI detector score as somewhat binary even though there are gray areas.

• Scores below 50%: human written
• Scores above 50%: AI written

In other words, if your score is below 50% (and for more safety, let's use 35%), it's likely okay to publish without risking a Google penalty for using AI.

You still need to use other SEO tactics to improve your rankings, like keyword usage, links, images, and other best practices.

More AI Detection Test Results

I'll post some more enlightening test results tomorrow morning (June 19, 2024). Watch my LinkedIn feed (follow me and click the bell to get notified).

If you've tested AI detectors, I'd love to hear about your results.

?? Like this? Please repost.

?? Follow me for daily executive and SEO coaching written by me, not AI.