A little more than a year after ChatGPT went live, certain predictions about how it will impact the Internet are starting to come true. That means AI spam is flooding the web.
Just last week, we had three examples of how this played out.
First, 404 Media, a new tech blog, wrote about how they had to change their website due to AI spam.
Recently, they noticed that AI-generated versions of the scoop were showing up on SEO-friendly spam sites and sometimes appearing above actual 404 Media articles in Google searches. Of course, the scammer makes money by placing ads on his AI-generated pages.
From 404 Media’s investigation into the cottage industry of article theft:
Over the past few weeks, Jason has been researching and experimenting with a series of AI tools that promise to “spin” articles for users. One is called SpinRewriter, which allows a user to create his 1,000 slightly different versions of the same article in one click and automatically publish them to as many of his WordPress sites as he wants using a paid plugin. Masu. It also provides tools that allow the user to manage as many of his websites as he wants from a single dashboard. a A company called Byword is gleefully promoting it. This strange trick (export your competitor’s sitemap, AI generates 1,800 article versions).
These AI-generated versions of stories hurt the news business, effectively taking clicks (and revenue) away from news organizations that spent time and money reporting in real time.
Next, Wired reported that The Hairpin, a popular indie blog from the 2010s, was taken over by AI click farmers, keeping some of the popular articles but replacing the names of the women who wrote them with men’s names. I wrote about the background. No way.
Finally, at the most pernicious end of the AI spam spectrum are AI-generated obituaries that are riddled with errors and cause great distress to grieving families. In 2021, long before ChatGPT, Wired reported that “obituary pirates” were scraping and copying the funeral home’s website for her. Now they are using his AI to create new and lucrative tactics such as creating YouTube videos and spam websites based on the information to capture search traffic from people looking for information about the recently deceased. is running.
The New York Times recently reported on the pain these AI-generated YouTube videos actually caused to grieving families. After a college student died after accidentally falling onto New York subway tracks, his YouTube videos and his AI-generated articles quickly appeared.
These AI-generated tips were in response to scammers noticing a spike in search interest for young people’s names and the word “subway.” The scammers quickly typed in these key terms, told the AI to write an article in a conversational tone, and pasted it onto their website, the Times reported. (Most of the details were wrong, but that didn’t stop the site from appearing in Google search.)
All three examples, 404 Media’s imitators, The Hairpin’s squatters, and obituary pirates, differ in details. But they all have one thing in common. That means bad actors, scammers, and spammers are using his AI to pump out tons of content and try to make money by showing up at the top of Google search results.
After all, this is not just a problem for journalists who have had their content stolen or families who are understandably saddened by digital grave robbers. This is a big problem for Google. In the end, users end up seeing garbage results, but they now have more and more attractive options for their searches, again thanks to her AI.
Google told the New York Times that it is aware of these spam articles and is working to address them (and has removed some because they violate its policies).
But bad actors are often one step ahead of the platform, like the AI-generated lewd images of Taylor Swift that went viral on X last week.
AI will fundamentally change the internet, for better or worse. It’s up to Google and the companies developing these AI tools to minimize the actual damage.
