Pod Insights Voices of Search // A Search Engine Optimization (SEO) & Content Marketing Podcast Google's View Of AI Content Revisited About this episode CEO at Originailty.ai, Jonathan Gillham, revisits Google's view of AI content. In this episode, Jonathan shares his perspectives on: exploring the impact of AI on SEO and content marketing and the role of detection technology in understanding AI's authenticity. Show NotesConnect With: Jonathan Gillham: Website // LinkedInThe Voices of Search Podcast: Email // LinkedIn // TwitterBenjamin Shapiro: Website // LinkedIn // TwitterSee Privacy Policy at https://art19.com/privacy and California Privacy Notice at https://art19.com/privacy#do-not-sell-my-info.
Transcript 0:00 Speaker 00 The Voices of Search Podcast is a proud member of the I Hear Everything Podcast Network. Looking to launch or scale your podcast? I Hear Everything delivers podcast production, growth, and monetization solutions that transform your words into profit. Ready to give your brand a voice? Then visit ihereeverything.com. Welcome to the Voices of Search Podcast, a member of the I Hear Everything Podcast Network. Ready to expedite your company's organic growth efforts? Sit back, relax, and get ready for your daily dose of search engine optimization wisdom. Here's today's host of the Voices of Search Podcast, Tyson Stockton.
0:43 Speaker 01 Hey, what's going on? My name's Tyson from previsible.io and joining me today is Jonathan Gilham, who is the CEO at originality.ai. Originality.ai uses AI to make the job of a copy editor 10 times more efficient, and they also let publishers hit publish with more integrity. Yesterday, John and I talked about content AI ethics and safety, and today we're going to continue the conversation by discussing Google's view of AI-generated content, and we're going to revisit that conversation since the last time we spoke. So with that, here's my conversation with John Gilham, CEO at originality.ai. John, welcome back.
1:27 Speaker 02 Yeah, thanks for having me again.
1:29 Speaker 01 So I know, I know we kind of, we had this conversation and I forget actually on the top of my head of when exactly it was. I think maybe it's like coming up on almost a year, nine months or so since we spoke last. But I wanted to kind of dive in and see, like pick your brain on what you're seeing, see if kind of like opinions have changed at all. Like, how is it evolving, you know, this idea of how Google's perceiving AI-generated content? So maybe like, we could first just kind of set the stage, like, what are you seeing from like a usage? Or is it kind of flatlined, going through the roof? Like, what's the status quo of the usage of AI content on the web?
2:14 Speaker 02 Yeah, so we can answer it from in terms of usage. Ahrefs had some some interesting data around sort of ninety seven percent of the web gets gets zero. Ninety seven percent of the content on the web gets zero clicks from Google. And then, so I can answer for the three percent that we see does get some clicks from from Google. So I would imagine in that ninety seven percent, there's a bunch of other stuff happening. But out of sort of the visibility within Google, so the top twenty results across a series of keywords, looking at sort of a hundred thousand data points. We're seeing about thirteen percent of the Google search results are filled with content that is likely AI.
2:58 Speaker 02 And that rose from sort of twenty twenty. It was when we looked at the study, it was like two percent, which is sort of right in line with our false positive amount. So that's probably mean that, you know, essentially none. And then it climbed after G P T three got released, starting increasing significantly with the G P T four release. And then what's been interesting is sort of since the Google March update, it's actually been accelerating, not not decelerating in terms of the amount of AI content that is showing up on the web. And so, yeah, we're out of right around thirteen fourteen percent right now.
3:31 Speaker 01 Wow, that's, I mean, I guess the first part of that, I feel like is like, okay, that's kind of what I would think from it. Like as technology advances, people's confidence in it, awareness kind of grows. So people start, you know, applying it more and more. And it's, you know, certainly something that didn't, you know, it's not like generated content was just from chat G P T. But like, it's really interesting with like that kind of March update. And I guess to that point, too, I want to make just like a little plug is you guys do have a study that, you know, showcases these numbers and gets into a bit more detail. So for the listeners, you can go over to the show notes and get links to that, which is on originality.ai's website.
4:16 Speaker 01 But with that, I want to go back to that piece of the March core update. So again, like makes sense, technology improves. What's kind of the deal then with the algo update and seeing a balloon, especially when Google's had kind of those mixed messages of, at least in the early days, like, don't use this.
4:36 Speaker 02 So also to walk walk through, get us kind of up to the March update timing from from like at least our lens. You know, Google first came out and said, hey, don't use AI and then came out and said like, ah, use, we don't, don't spam. However, you create content, we'll, we'll figure out whether it's spam or not. But we're not saying whether it's created by humans or by AI. Spam is bad. Good content is good. It's kind of what they said. And it like that, that wording makes sense from a company that is trying to ensure that they don't have their codec moment where they get sort of put out of business by the technology that they invented. And they need to be this sort of like AI forward company. But it doesn't sort of, I think if you ask anyone, do you want your search results filled with nothing but AI content? The answer is no, I just
5:21 Speaker 02 go to the, I just go to the, if I, if I know all I'm getting is AI content, that I'm just going to go to the, the AI that knows more about me and, and ask them the question. And so that sort of Google's facing this existential threat. And then kind of with the March update, they'd seen this sort of wave of AI coming and increasing in their search results. The update did a few things. I call it a bit of a psyop, where they did a significant manual action at the start of the update, where they de-indexed a ton of sites. We looked at 70,000 sites, about 2% to 3% of those sites got de-indexed out of that sample. That sample was a, were decent quality sites as well.
6:05 Speaker 02 These are sites that were on Mediavine or AdThrive, Raptive, but so on, sort of reasonable quality sites, had to have some, some decent traffic numbers to get on those platforms. And about 2% to 3% of them got, got totally, totally wiped out. I call it sort of a psyops, because they did the, the, almost the most marketing push that they had ever seen around them, saying like, this update's going to deal with 40% of, of spam on the internet. And then on the eve of the update, they took these manual actions against the largest number of sites that I'd ever seen sort of a manual action get applied to in one consistent wave. So it was this kind of coordinated effort to try and send this. Signal of don't spam us and if you do, we're going to find you.
6:49 Speaker 02 Following that with like, so again, like to your point, like that kind of all makes sense. But then why has it been this update? Why, why has there been this increase in amount of AI content in the search results after? And I think what that says is that Google maybe won the battle with AI content overrunning their search results. But I think they're struggling to win the war. And there, I mean, the amount of content that is continuing to show up in their search results is concerning to them. And they have not yet found a way to efficiently and algorithmically deal with it. And so they rely on these sort of manual action psyop efforts to, to deal with it.
7:28 Speaker 01 That's super interesting and I, I feel like that, that piece, at least kind of like in the SEO news arena, kind of got totally missed where it's like, we know and like most of the things that were discussed about it was, hey, this is like one of the longest running updates at nineteen days. Helpful contents baked into the core and kind of like, you know, similar narratives that we've heard. Maybe I completely missed it, but it's like, this is the first time that I've heard that on the eve of the update, there was the manual penalties from that. And I, I want to get into like the rise of it, but like, I guess staying with just the manual penalties.
8:08 Speaker 01 From what you saw as far as like the types of websites, like, did you notice or see like any elements or any kind of like common themes with the types or like volumes of content? Like, do you have any, any ideas of kind of what was the, the focus here?
8:29 Speaker 02 Yeah. So mass publishing of spam content was, was obviously the, the commonality. And I think in, in twenty twenty four, not all AI content is spam, but anybody that's spamming the internet in twenty twenty four is, is using AI. So that's sort of the result of that. So what was identified? Mass publishing of content was, was the first sort of obvious. And then when we went and looked at the type of content, the majority of sites had the majority of their content being AI generated as content. Now, again, that's not all AI content is spam, but all spam is AI. And so it made sense to see that it was, if somebody published a thousand articles in a month, unless they're a big publishing company, they were probably leveraging AI to, to do so.
9:15 Speaker 01 Interesting. And so I mean, I guess, obviously, there's like various ways of kind of like flagging that, whether it's like velocity of publishing and whatever else in between. And it kind of feels like it also like fits within that narrative that we've been hearing more often, which is, you know, content that's generated just to rank high in search engines is content that's going to be done. You just didn't expect, or at least I didn't expect that that was going to be in the form of like a manual panel versus, you know, as part of the update.
9:48 Speaker 02 Agreed. It was, it's a, you know, I, I hope, I hope somebody leaked something at some point about the sort of inner narrative that occurred at Google around those manual actions and the timing of it. So just the manual actions are so disconnected in terms of the action that was taken from the update, yet it happened on the, like literally the eve of the update. And so, yeah, kind of why, why I like to call it a bit of a psyops update.
10:12 Speaker 01 Yeah, it kind of just gets lumped into buried in this larger shift or like variance, which I mean, even after we've been seeing, you know, quite a bit of, you know, fluctuation in search results, like even post completion of the update. But now I, I kind of want to like dive in a little bit more on this, like this rise that you guys are showing in this study. And I'm kind of looking at the graph that you're showing here and, you know, from 2020 going up until kind of really the, the March core update, like there's fluctuations, but it's like, it's a pretty steady, consistent growth trend.
10:54 Speaker 01 And then it hits that point, jumps up, and it looks like it's not slowing down either. Like, you know, it's not like it hit the update and then plateaued and it's just kind of like trailing off. Like, yeah, it's not, you know, the exact same growth every single kind of like interval, but it still appears to be like increasing at a, at a higher velocity. Is this from your, your kind of perspective is like, yeah, like cats out of the bag. There's no way of being like, hey, we're not gonna allow this AI content because clearly they can identify it. But it's like, it's just the usage is so great that you kind
11:35 Speaker 01 of just have to live with it? Or like, I mean, maybe that's a little hypothesis driven, but like, what's your perspective on it?
11:43 Speaker 02 So hard to not sort of apply my own lens to it in terms of being an AI detection provider. So the largest, I view these updates, sort of like call them like as individual battles that Google is having with this, with this sort of AI spam that they are needing to fight. The largest drop in the AI content that we saw in search results occurred at the March 4th. The largest relative drop occurred where the graph was going up fairly smooth. And then it had its biggest drop at that March update. So the, out of any month, Google nuked the most AI content at the time of that March update. But then it has continued, it has continued on this sort of growing curve since then.
12:25 Speaker 02 It's still only at like 13%, so it's still sort of relative or 14%. So still kind of relatively small to all of the search results, but accelerated. And so I guess I'd say that, that I think this growth represents what is happening in the world with the number of people that are using AI content. The sort of caveat that I throw into it is that this tells me that Google didn't, doesn't like it. That Google is, will go to these big efforts to try and deal with it, but are struggling to do so. And so sort of won the war or won the battle on that sort of like March 5th update around sort of nuking a ton of AI content. But it's just being overwhelmed with the amount and have not been capable yet to algorithmically
13:10 Speaker 02 deal with it and are probably tracking the same kind of graph or something similar. But are probably not very happy because at what point, if this, if this graph went to 100% and their search results are filled with nothing but content that had been AI generated, I don't think Google would be as valuable a company as they are right now.
13:29 Speaker 03 Time for a one minute break to hear from our sponsor, Previsible. So you're looking for SEO help and you got a couple of options. You could start replying to spam from agencies that claim they can get you to rank number one on Google. You can pay an hourly rate for a consultant who will inevitably nickel and dime you with hourly charges. Or you can work with a cookie cutter agency to quickly launch a strategy less project with low success. None of those sound very good. Now, do they? Well, that's where Pre-Visible's integrated consulting model comes in. Pre-Visible draws from a collective 40 years of SEO and digital marketing experience to unlock your organic growth opportunities.
14:06 Speaker 03 They build custom solutions that combine strategy, technical expertise, content, and reporting to effectively operationalize SEO for your business. Pre-Visible's four-stage approach ensures that your SEO programs thrive by starting off with a strategy-first approach. Then they support you in your efforts to create quality content, help you identify technical issues, and most importantly, they'll work with your cross-functional teams to integrate your SEO strategies to make sure that your SEO budget actually drives results, not just your agency's bottom line. So join brands like Yelp, eBay, Canva, Atlassian, Square, all who rely on the SEO consultants at Pre-Visible.
14:46 Speaker 03 For more information, go to previsible.io. That's pre-visible, P-R-E-V-I-S-I-B-L-E dot I-O.
14:58 Speaker 01 That's super interesting. One other piece that I want to touch on is like within your study here too. You're showing, you know, not just like the volume of AI-generated content showing up in, you know, like the top rankings, but you're also showing kind of like the position in which that it's ranking. And looking at the graph that you have, like, everything is like reasonably tight clustered and it's like fairly steady in that. And then all of a sudden, also around that March update time period, you see like a bit more kind of, I don't know, it's like a clear change in that point.
15:43 Speaker 01 Like, have you noticed anything else besides kind of volume of content that's showing up?
15:52 Speaker 02 So exactly, the graph got really noisy after that update. So it's showing that, and I think this this tracks with sort of the the thesis around the the merge update and the helpful content update. But started using a wide range of signals for trying to determine what was helpful and not helpful. And so I think this this just sort of represents the noisiness that is existing. And there's a pretty tight clustering of sort of type of factors that were involved in determining what content would rank. And then I think as as things have progressed, the number of factors that that are used in identifying what content ranks
16:34 Speaker 02 has continued to grow within within Google and the complexity has grown and therefore the noise has has also grown. So I think what I think that's what I I'm still confused by that by that graph that shows this sort of like significant range of what position content that is likely AI was ranking, it massively got noisy post the merge update. And I don't I don't have a great answer yet. So it is it's a very interesting graph and I'm I'd love to hear anyone's thoughts on it.
17:03 Speaker 01 Yeah, and I mean, I think like noisy is at least like for, you know, the period of time that we have right now and looking at is probably like the best description that I could give to it, where it's like you do see a few outliers where it's like top ranking positions are kind of like jumping up in the percentages of websites. But then those also seem to be kind of like one offs, highly fluctuation. And it's like, maybe it's going out on a limb here, but it's like the only sense that I think I could make from that is there is a ton of noise going on and it could also relate to
17:44 Speaker 01 the amount of volatility that's been detected in the SERPs since the March. And, you know, just three or four weeks back or so, I had Barry Schwartz on and he was talking about like the volatility that we've been seeing post, you know, the close or the official close of March core updates complete. And so like, maybe these fluctuations could also at least have like a correlation to that same like turbulence. But yeah, it's it's it's certainly like an interesting piece here. And so I think like, bring you back around. Last time we were talking, it wasn't a clear Google likes or doesn't like it. From some of these studies
18:27 Speaker 01 that you're showing, I think there's, you know, at least from my perspective, what seems to be some clear signals that there's things that Google is definitely doesn't like. And there's things that Google definitely like, seems to be, I don't know, call it scared or like, very cautious of. But like, it seems like they're also with like, the rise of it is, you know, it's it's also hasn't changed in some sense, like the the story's gotten a little more interesting, a little more complex. But it's still not a clear yes or no, good or bad, like AI content.
19:05 Speaker 02 Agree. It the story's definitely gotten more complex. But I think some of the, I think where what where it has evolved to now is I think we can make some clear, some clear answers on a few things. And some of them unfortunately aren't going to be that obvious, or aren't going to be that insightful. But it's like, Google doesn't like mass publishing AI generated spam. That's that one, I think, has become incredibly clear. They take, you know, not not just here, but like, you know, the the SEO heist tweet site getting nuked. You know, I think there's lots of examples of somebody going and saying, hey, I can hit this one button and publish an outrageous number of posts all generated by AI, and I'm going to crush it in Google. I think that end of the spectrum is the answer has become incredibly clear. Bad. Hate Google hates it. And then I
19:50 Speaker 02 would also say that I think it has also become clear that publishing AI generated content is riskier than publishing human generated content. I think that people could could argue with that, and I think that would be a conversation worth having. But I my, I think the data is saying that if you have a site that is worth a ton of money, ton of value to you, You can choose to publish AI-generated content or human-generated content. It is riskier to publish AI-generated content. I would say that that answer has become clear, at least to my eyes, but I know people would. Again, it would be worth having a conversation if someone wanted to argue with that because there would be valuable points to the opposite side.
20:30 Speaker 02 Yeah, so I said, that's like on the on the spectrum of of kind of where the conclusions that I think are. We can reasonably make at this stage that maybe couldn't have made, I mean, could have assumed at this kind of moment, last time we chatted at the end of last year. Could assume that's what was going to be happening. I think we've seen data that supports AI-generated spam, mass publishing bad. AI content riskier than human content.
20:54 Speaker 01 I totally agree with those and I feel like really this last graph that we are referencing here, seeing like the volatility of the AI-generated content that's ranking would kind of lean into that second point. And I feel like, like, yeah, you could make the argument that it's like, yeah, if I am very careful and very deliberate in how I'm deploying AI content. Like, I'm taking a calculated risk, but I'm also, you know, measuring that risk in my decisions. And it would be like, kind of similar to, you know, let's say, like going skiing or snowboarding. It's like, or maybe a better example would be like choosing like riding a motorcycle.
21:38 Speaker 01 Like, you could say riding a motorcycle is more dangerous than driving a car. But it's like you can also ride a motorcycle in a safe way, but it's like for the like larger conversation. There is that factor that maybe you can't control. And so it's like, yeah, motorcycle is definitely riskier. And so if you're wanting to hedge on that safety side, it's like, yeah, you're going to pick the car in that scenario.
22:04 Speaker 02 Yeah, you know, throwing the baby on the back of the motorcycle.
22:07 Speaker 01 Fair. At least you'd hope not.
22:10 Speaker 02 Unless, depends what country you're in, I guess. So there's definitely, definitely some countries where that's well, well-established mode of transportation.
22:18 Speaker 01 I mean, I don't know, for me, like, just like a couple of months ago, like, I was in a in a state that didn't have helmet laws. And being from California, I was like, just mind blown. I was like, how is this a thing? Like, someone's riding a motorcycle without a helmet? Wisconsin, by the way.
22:34 Speaker 02 Yeah, we're Canadian. And so, yeah, one of those, one of those things when you go to some states where it's like, that's, that's shocking that, that there's, there's no helmet being worn. Yeah, all, all to their own.
22:48 Speaker 01 Fair enough. Well, one, just really enjoyed the conversation and kind of the, also the numbers that you're helping kind of put behind this narrative because, you know, it is, you know, sometimes very anecdotal of like personal experiences. So I think this is super helpful. Anything else you'd want to leave with the listeners on the topic?
23:10 Speaker 02 Yeah, it's fascinating time. I think SEOs have always been most successful when, when sort of using data and math and logic. And I think the, you know, AI content, AI detection is no different that I, I think we'll, we'll eventually work through and have some, some common consensus on what's the right state of play. But I'm, yeah, sort of definitely here, here for that journey and that conversation and throwing as much data into it as, as we possibly can.
23:37 Speaker 01 Absolutely. Well, with that, that wraps up this episode of the Voice of Search Podcast. Thanks again to John Gilman from originality.ai for joining us. It's always a pleasure to have you on. And if you'd like to get in touch with John, you can find a link to his LinkedIn profile in the show notes, or go on over to his company's website at originality.ai. And also be sure to check the show notes for the two different studies that we referenced in this episode.
24:05 Speaker 03 Okay. Thanks to Tyson Stockton, our guest host. If you'd like to get in touch with Tyson, you could find a link to his LinkedIn profile in our show notes. You can contact him on Twitter where his handle is Tyson underscore Stockton. Or if your team is interested in SEO consulting or organizational education, you can always head to their company's website, which is previsible.io. That's P R E V I S I B L E dot I O. Just one more link in our show notes I'd like to tell you about. If you didn't have a chance to take notes while you were listening to this podcast, head over to voicesofsearch.com where we have summaries of all of our episodes and contact information for our guests. You can also subscribe to our weekly newsletter and you can even send us your topic suggestions or your marketing questions, which we'll answer live on our
24:48 Speaker 03 show. Of course, you can always reach out on social media. Our handle is @voicesofsearch on LinkedIn, Twitter, Instagram, Facebook, or you can contact me directly. My handle is @benjschap, B-E-N-J-S-H-A-P. And if you haven't subscribed yet and you want a daily stream of SEO and content marketing insights in your podcast feed, we're going to publish an episode every day during the work week. So hit that subscribe button in your podcast app and we'll be back in your feed tomorrow morning. All right, that's it for today. But until next time, remember, the answers are always in the data.
Previous Content AI Ethics & Safety Next How To Be More Strategic With Your SEO Page updated November 15, 2026 · transcript and summary are machine-generated