Google doesn’t ban AI content, so why are we obsessing over detector scores?

Neylo

Neylo

Power Member
Jr. VIP
Joined
Jun 18, 2012
Messages
549
Reaction score
408
Google doesn’t care about AI text just because it’s AI text.

I keep seeing people run an article through a detector, get a high score, and immediately assume Google won’t rank it.

But how did we get from “ZeroGPT thinks this is AI” to “Google will penalise this”?

Those are two completely different claims.

Google has already explained its position. Appropriate use of AI isn’t against its guidelines. Generating loads of pages without adding anything useful can be a problem, but using AI doesn’t automatically put your content in that category.

You can read it straight from Google here: Google’s guidance on AI content.

some examples of pages ranking 1st on google:

Keyword: benefits of cloud computing
ZeroGPT score: 66.3%
Page: IBM’s cloud computing article

Keyword: benefits of workflow automation
ZeroGPT score: 96.8%
Page: Celigo’s workflow automation article

keyword: how does an ultrasonic cleaner work
https://www.besttechnologyinc.com/ultrasonic-cleaning-systems/how-do-ultrasonics-work/
85.4%

how does a hydrogen water bottle work
https://ocemida.com/blogs/hydrogen-water/how-does-a-hydrogen-water-bottle-work
88.8%

how does a cold plunge chiller work
https://plungecrafters.com/blogs/diy-cold-plunge-chronicles/how-cold-plunge-chillers-work
84.2%


I could go on and on....


If ZeroGPT is right, content it considers AI-generated is sitting at #1 and #4.

If ZeroGPT is wrong, why would you use its verdict as a reason to rewrite a perfectly useful article?

Either way, the score doesn’t tell you whether Google will rank the page.

“But Google can detect AI”

Maybe it can detect certain patterns. Maybe it uses signals we don’t know about. We don’t have the full picture.

But even identifying AI involvement wouldn’t tell us what Google does with that information.

Being able to recognise something doesn’t mean automatically penalising it.

And there’s no reason to assume a commercial detector sees a page the same way Google does. Getting 96.8% on ZeroGPT doesn’t mean Google has given your article a matching score somewhere.
 
Advertise on BHW
you would think that but I ran tests on 20 different websites, all with AI content targetting EXTREMELY easy keywords that maybe only 3 or 4 other websites targetted them.

All 20 started getting some clicks from google in the first month.
By the 2nd month the impressions completely flatlines for 18 of the websites.
Only 2 websites survived and are thriving to this day (1 year later). I have not added anymore content.

I actually even built some guest posts (from real websites in the niche) for all of the websites.

And it still didn't help. The websites had no reason to be penalized as they all had only ~30-60 articles with just a handful of links which were fully legit and high quality.

The articles were fully made with AI but also ran a lot of prompting engineering to change the way they read so they don't sound AI at all and have no obvious AI footprints.

Yet all 18 flatlined all around the same time.

Why the other 2 survived I have no clue.

It's well known that Claude, ChatGPT etc all have hidden watermarks placed on text, pretty sure it's easy for something like google to check those watermarks.
 
Did those 18 sites lose rankings, or were their pages actually deindexed? Also, did GSC show any manual actions? A drop in impressions alone doesn't prove Google penalized them for using AI content. There could be other reasons behind the traffic loss.
 
I don't bother with detectors. Waste of time for me. English is my second language, and I've had my own writing flagged as AI after correcting the grammar lol. That score isn't Google's verdict; allowing AI doesn't mean allowing spam.
I've had good success with AI blogs and tiered link building. I use AI for writing now rather than outsourcing, and tbh I can't remember when I last paid a native writer. I'd spend that time checking facts and making the page useful, not rewriting it to please a detector.
 
The key distinction is between detecting how content was produced and evaluating whether that content is useful. A detector score cannot tell you whether an article answers the searcher's question accurately.

I would spend more time checking factual claims, original examples and whether the article adds something missing from competing pages. The example of 18 sites losing impressions is interesting, but I would want to compare indexing status, content overlap and search demand before attributing the decline to AI detection.
 
Neylo said:
Google doesn’t care about AI text just because it’s AI text.

I keep seeing people run an article through a detector, get a high score, and immediately assume Google won’t rank it.

But how did we get from “ZeroGPT thinks this is AI” to “Google will penalise this”?

Those are two completely different claims.

Google has already explained its position. Appropriate use of AI isn’t against its guidelines. Generating loads of pages without adding anything useful can be a problem, but using AI doesn’t automatically put your content in that category.

You can read it straight from Google here: https://developers.google.com/search/blog/2023/02/google-search-and-ai-content.

some examples of pages ranking 1st on google:

Keyword: benefits of cloud computing
ZeroGPT score: 66.3%
Page: https://www.ibm.com/think/topics/cloud-computing-benefits

Keyword: benefits of workflow automation
ZeroGPT score: 96.8%
Page: https://www.celigo.com/blog/workflow-automation/

keyword: how does an ultrasonic cleaner work
https://www.besttechnologyinc.com/ultrasonic-cleaning-systems/how-do-ultrasonics-work/
85.4%

how does a hydrogen water bottle work
https://ocemida.com/blogs/hydrogen-water/how-does-a-hydrogen-water-bottle-work
88.8%

how does a cold plunge chiller work
https://plungecrafters.com/blogs/diy-cold-plunge-chronicles/how-cold-plunge-chillers-work
84.2%


I could go on and on....


If ZeroGPT is right, content it considers AI-generated is sitting at #1 and #4.

If ZeroGPT is wrong, why would you use its verdict as a reason to rewrite a perfectly useful article?

Either way, the score doesn’t tell you whether Google will rank the page.

“But Google can detect AI”

Maybe it can detect certain patterns. Maybe it uses signals we don’t know about. We don’t have the full picture.

But even identifying AI involvement wouldn’t tell us what Google does with that information.

Being able to recognise something doesn’t mean automatically penalising it.

And there’s no reason to assume a commercial detector sees a page the same way Google does. Getting 96.8% on ZeroGPT doesn’t mean Google has given your article a matching score somewhere.
Click to expand...
what I have noticed is that attempting to achieve a low score in ai detector takes time yet cannot assure the performance of the content in the search engines. I would prefer using the time checking for click impressions and relevance to the current ranks rather than wasting my time with the ai detector. with the 18 site have you observed any pattern in the niche or the content before the drop in impressions?
 
Neylo said:
Google doesn’t care about AI text just because it’s AI text.

I keep seeing people run an article through a detector, get a high score, and immediately assume Google won’t rank it.

But how did we get from “ZeroGPT thinks this is AI” to “Google will penalise this”?

Those are two completely different claims.

Google has already explained its position. Appropriate use of AI isn’t against its guidelines. Generating loads of pages without adding anything useful can be a problem, but using AI doesn’t automatically put your content in that category.

You can read it straight from Google here: Google’s guidance on AI content.

some examples of pages ranking 1st on google:

Keyword: benefits of cloud computing
ZeroGPT score: 66.3%
Page: IBM’s cloud computing article

Keyword: benefits of workflow automation
ZeroGPT score: 96.8%
Page: Celigo’s workflow automation article

keyword: how does an ultrasonic cleaner work
https://www.besttechnologyinc.com/ultrasonic-cleaning-systems/how-do-ultrasonics-work/
85.4%

how does a hydrogen water bottle work
https://ocemida.com/blogs/hydrogen-water/how-does-a-hydrogen-water-bottle-work
88.8%

how does a cold plunge chiller work
https://plungecrafters.com/blogs/diy-cold-plunge-chronicles/how-cold-plunge-chillers-work
84.2%


I could go on and on....


If ZeroGPT is right, content it considers AI-generated is sitting at #1 and #4.

If ZeroGPT is wrong, why would you use its verdict as a reason to rewrite a perfectly useful article?

Either way, the score doesn’t tell you whether Google will rank the page.

“But Google can detect AI”

Maybe it can detect certain patterns. Maybe it uses signals we don’t know about. We don’t have the full picture.

But even identifying AI involvement wouldn’t tell us what Google does with that information.

Being able to recognise something doesn’t mean automatically penalising it.

And there’s no reason to assume a commercial detector sees a page the same way Google does. Getting 96.8% on ZeroGPT doesn’t mean Google has given your article a matching score somewhere.
Click to expand...
imo detector scores are now more distracting than useful.i have seen people spending more effort on making an article appear more human to the tool rather than enhancing the content itself.ultimately, the response that one receives from readers is far more important than any dettector score.
 
Neylo said:
Google doesn’t care about AI text just because it’s AI text.

I keep seeing people run an article through a detector, get a high score, and immediately assume Google won’t rank it.

But how did we get from “ZeroGPT thinks this is AI” to “Google will penalise this”?

Those are two completely different claims.

Google has already explained its position. Appropriate use of AI isn’t against its guidelines. Generating loads of pages without adding anything useful can be a problem, but using AI doesn’t automatically put your content in that category.

You can read it straight from Google here: Google’s guidance on AI content.

some examples of pages ranking 1st on google:

Keyword: benefits of cloud computing
ZeroGPT score: 66.3%
Page: IBM’s cloud computing article

Keyword: benefits of workflow automation
ZeroGPT score: 96.8%
Page: Celigo’s workflow automation article

keyword: how does an ultrasonic cleaner work
https://www.besttechnologyinc.com/ultrasonic-cleaning-systems/how-do-ultrasonics-work/
85.4%

how does a hydrogen water bottle work
https://ocemida.com/blogs/hydrogen-water/how-does-a-hydrogen-water-bottle-work
88.8%

how does a cold plunge chiller work
https://plungecrafters.com/blogs/diy-cold-plunge-chronicles/how-cold-plunge-chillers-work
84.2%


I could go on and on....


If ZeroGPT is right, content it considers AI-generated is sitting at #1 and #4.

If ZeroGPT is wrong, why would you use its verdict as a reason to rewrite a perfectly useful article?

Either way, the score doesn’t tell you whether Google will rank the page.

“But Google can detect AI”

Maybe it can detect certain patterns. Maybe it uses signals we don’t know about. We don’t have the full picture.

But even identifying AI involvement wouldn’t tell us what Google does with that information.

Being able to recognise something doesn’t mean automatically penalising it.

And there’s no reason to assume a commercial detector sees a page the same way Google does. Getting 96.8% on ZeroGPT doesn’t mean Google has given your article a matching score somewhere.
Click to expand...
This has been known for a while, but for me it's not just about Google alone, sites being monetized with display Ads do get suspended sometimes due to AI content.
 
I don't think you need to worry about detector scores from a ranking standpoint.

Instead, I think you should worry about AI content from a conversion point of view.

If I see em dashes and clear AI content, I'm likely leaving the page because I have no intent reading it anymore.
 
AI detection scores are not a reliable way to measure content quality or predict rankings.

I've seen plenty of AI assisted content perform well, the real focus should be whether the article answers the search intent, provides useful information, and offers something valuable to readers.
 
Neylo said:
Google doesn’t care about AI text just because it’s AI text.

I keep seeing people run an article through a detector, get a high score, and immediately assume Google won’t rank it.

But how did we get from “ZeroGPT thinks this is AI” to “Google will penalise this”?

Those are two completely different claims.

Google has already explained its position. Appropriate use of AI isn’t against its guidelines. Generating loads of pages without adding anything useful can be a problem, but using AI doesn’t automatically put your content in that category.

You can read it straight from Google here: Google’s guidance on AI content.

some examples of pages ranking 1st on google:

Keyword: benefits of cloud computing
ZeroGPT score: 66.3%
Page: IBM’s cloud computing article

Keyword: benefits of workflow automation
ZeroGPT score: 96.8%
Page: Celigo’s workflow automation article

keyword: how does an ultrasonic cleaner work
https://www.besttechnologyinc.com/ultrasonic-cleaning-systems/how-do-ultrasonics-work/
85.4%

how does a hydrogen water bottle work
https://ocemida.com/blogs/hydrogen-water/how-does-a-hydrogen-water-bottle-work
88.8%

how does a cold plunge chiller work
https://plungecrafters.com/blogs/diy-cold-plunge-chronicles/how-cold-plunge-chillers-work
84.2%


I could go on and on....


If ZeroGPT is right, content it considers AI-generated is sitting at #1 and #4.

If ZeroGPT is wrong, why would you use its verdict as a reason to rewrite a perfectly useful article?

Either way, the score doesn’t tell you whether Google will rank the page.

“But Google can detect AI”

Maybe it can detect certain patterns. Maybe it uses signals we don’t know about. We don’t have the full picture.

But even identifying AI involvement wouldn’t tell us what Google does with that information.

Being able to recognise something doesn’t mean automatically penalising it.

And there’s no reason to assume a commercial detector sees a page the same way Google does. Getting 96.8% on ZeroGPT doesn’t mean Google has given your article a matching score somewhere.
Click to expand...
Tbh, when I say something about this problem, I think that email verification is not the whole thing, as there is always a chance that the owner of the email will never receive it. for me, it would be much more interesting to know if this list works well for any industry, and to see if some demographics work better with this list than others. do you track the response rate nd bounce or do you track the nu of emails sent?
 
On the one hand, I agree that google actually doesn't care too much about AI generated text. Moreover, I figured out Gemini often works better with AI articles.
On the other hand, AI generated articles has less ability to collect clicks. It's not friendly for organic traffic, to my mind
 
Neylo said:
Google doesn’t care about AI text just because it’s AI text.

I keep seeing people run an article through a detector, get a high score, and immediately assume Google won’t rank it.

But how did we get from “ZeroGPT thinks this is AI” to “Google will penalise this”?

Those are two completely different claims.

Google has already explained its position. Appropriate use of AI isn’t against its guidelines. Generating loads of pages without adding anything useful can be a problem, but using AI doesn’t automatically put your content in that category.

You can read it straight from Google here: https://developers.google.com/search/blog/2023/02/google-search-and-ai-content.

some examples of pages ranking 1st on google:

Keyword: benefits of cloud computing
ZeroGPT score: 66.3%
Page: https://www.ibm.com/think/topics/cloud-computing-benefits

Keyword: benefits of workflow automation
ZeroGPT score: 96.8%
Page: https://www.celigo.com/blog/workflow-automation/

keyword: how does an ultrasonic cleaner work
https://www.besttechnologyinc.com/ultrasonic-cleaning-systems/how-do-ultrasonics-work/
85.4%

how does a hydrogen water bottle work
https://ocemida.com/blogs/hydrogen-water/how-does-a-hydrogen-water-bottle-work
88.8%

how does a cold plunge chiller work
https://plungecrafters.com/blogs/diy-cold-plunge-chronicles/how-cold-plunge-chillers-work
84.2%


I could go on and on....


If ZeroGPT is right, content it considers AI-generated is sitting at #1 and #4.

If ZeroGPT is wrong, why would you use its verdict as a reason to rewrite a perfectly useful article?

Either way, the score doesn’t tell you whether Google will rank the page.

“But Google can detect AI”

Maybe it can detect certain patterns. Maybe it uses signals we don’t know about. We don’t have the full picture.

But even identifying AI involvement wouldn’t tell us what Google does with that information.

Being able to recognise something doesn’t mean automatically penalising it.

And there’s no reason to assume a commercial detector sees a page the same way Google does. Getting 96.8% on ZeroGPT doesn’t mean Google has given your article a matching score somewhere.
Click to expand...
honestly speaking, I do not really care much about the detection scores since there have been a lot of genuine articles labeled as AI generated before. for me, the thing that matters most is whether or not the article answers people queries and holds their interest in the website. Rankings are alright, but if people visit the website nd leave straight away then definitely there is something wrong. from my perspective, true statistics say more than detector scores.
 
Advertise on BHW
You must log in or register to reply here.
Sign up now!

Main Menu

Home Main Forum List Black Hat SEO White Hat SEO BHW Newbie Guide Blogging Black Hat Tools Social Media Downloads

Marketplace

Account Selling Content / Copywriting EBooks / Methods Hosting Misc Proxies For Sale SEO - Link building SEO - Packages Social Media Social Media - Panels Web Design More

Making Money

Affiliate Programs Hire a Freelancer IM Journeys Making Money Pay Per Click Site Flipping Want to Buy

BlackHat World

Dispute Resolution Forum Suggestions & Feedback Introductions Lounge
Back
Top