Google is starting to squash more spam and AI in search results
![A Google logo sits at the center of ominous concentric circles](https://cdn.vox-cdn.com/thumbor/NhLQOIjjKA9QsV8Rx7HIeiSkwyw=/0x0:2040x1360/1310x873/cdn.vox-cdn.com/uploads/chorus_image/image/73183869/acastro_180427_1777_0001.0.jpg)
Google is rolling out a few new changes to its ranking systems in search, which are designed to help surface good content in your results and hide some of the worst and most cynical stuff on the web. The company says that it is doing a better job of downranking content that exists only to summarize other content — which can sometimes be normal SEO stuff but is also increasingly a job for generative AI tools — and in combatting some of the tricks people use to trick its ranking systems.
There are always people trying to manipulate their way to the top of Google results. That’s just a fact of the web and a fact of life for Google’s search teams. Google is always making changes to its ranking algorithms, too, in an effort to improve search results. We never hear about most of those changes. “You only see the ones that sort of slipped by the controls, as it were,” says Pandu Nayak, a VP of search at Google. “Unfortunately, these are not things you can just wave a magic wand and get rid of.”
For Google to announce the changes it’s making signals two things. First, that these are big changes that could meaningfully change your search experience — Nayak says that Google’s measurements show a reduction in “unhelpful content” by up to 40 percent. And second, that Google is sending a message to the web: your spammy, sketchy behavior ends now.
Nayak lays out three examples of what Google now considers spammy behavior and intends to downrank. The first is content at scale: the sites that create thousands of low-quality articles a day, either through low-paid contractors or AI generators, and target that content at search results. Nayak points to obituary spam — which The Verge’s Mia Sato recently wrote about — as an example of a problem to be solved here.
The second spammy behavior is what Nayak calls “site reputation abuse.” This is when an otherwise respectable website rents out part of its site for spammy nonsense; I won’t name and shame anyone here, but you’ve surely seen the sites that make you wonder why they have coupons or why there’s a whole part of the site that seems irrelevant and AI-generated. The third is “expired domain abuse,” which is when someone buys an abandoned but high-ranking domain and fills it with crummy content that then jumps to the top of search. The current state of The Hairpin is one example of how this can happen, which Wired has covered well in recent weeks.
For those engaging in site reputation abuse, Nayak says Google is giving the sites 60 days to cut it out before it makes the ranking changes. The others go into effect now. Google has a spam problem, it knows it, and it’s trying to shut it down. “The healthy, high-quality ecosystem is exactly the one that gets affected when spammers and low-quality purveyors of information get control of ranking,” Nayak says.
The job is not done, of course. The reckoning over AI-generated content — what it means, who wants it, how it should rank — is only just beginning and will cause Google plenty of internal headaches as it both tries to bring AI to everyone and tries to save the web from being overrun by it. (Even Google’s own search engine is increasingly an AI machine.) And there will always be new, sneakier ways to game your way to the top of search results. This is a headache of Google’s own making: most of the chum on the web exists entirely to game Google, and so Google will always be one step behind.
But for Google to continue to be Google, it has to be good at finding good stuff on the web. The company has been signaling for a while that it plans to care about and prioritize humans over machines and real content over clickbait, and it’s beginning to make moves in that direction. But it’s a long road ahead.