Back in April I posted a blog titled Dark Ages 2.0. While Dark Ages 2.0 is certainly worth a complete read, the main point was simply that AI is probably not going to lead to a utopian panacea of content creation and information dissemination. Nor is it going to be the end of humanity as we know it. What it is likely going to be used for most is to make a handful of corporations a crap-ton of money by creating massive amounts of content designed solely to capture individual’s attention and advertising dollars.
Less than a month after that article was posted, we now have a real-world example of what Phase One of this is going to look like.
As you may be aware, I do quite a bit of writing, and appear regularly on TV and radio to render commentary on a wide range of issues that are specifically related to Branding, Messaging, and Marketing. I’m often called to provide insights regarding the news of the day.
“How damaging is the Stormy Daniels Testimony to the Trump ‘brand’?”
“Will there be fallout for businesses that either publicly stand with Israel or publicly support Palestine?”
“Will Bud Light ever recover?”
These are just a smattering of recent topics I’ve weighed in on across several media platforms. Doing these media hits, I also generate a fair share of content, much of which is available on our website m.network. I’m not the only one who creates content for our website. Most of our employees and occasionally some of our clients also contribute. We do this, because we obviously love sharing our thoughts, but there are two more important reasons why we do this.
- We know that having fresh content posted to a website is one of many things that search engine algorithms like, and thus, it increases our visibility.
- We like to experiment with content on our site so we can advise our clients what works. We can also tell them what doesn’t work. In other words, we prefer to fail on our site rather than our clients’.
And that’s really what this post is about – us failing so you don’t. Here’s what happened.
Not long ago, my name popped up in a Google alert. When I looked at the headline I was, as you could imagine, a bit confused:
Evidently, I had gotten a new job!
And the write-up of me, well… let me just say, I couldn’t have done better myself!
Also, I couldn’t help but notice that all the fine content that this new creative agency had on their website was…
… OUR F$%@!-ING CONTENT!!!
This was our stuff. These were our clients. They took the article about one of our former interns!! They took the article our Design Director wrote about his service dog. They even used the dates the articles were posted to our website as the dates the articles were posted to their website. As if they’d been in business for years. What the…
Except, it wasn’t quite our content. It was an oddly watered down form our content. It was our topics and a couple of facts from our stuff, but it wasn’t what we wrote. Opening some of the articles made it clear exactly what was going on. Our website had been scraped. Evidently this is a growing problem and one that has almost no defense. Scraping is what happens when content from one website is fed into an AI platform like ChatGPT or Bard and used as source material.
It’s clear that this is exactly what happened to us. The phony site was created by some AI platform scraping the content from m.network. Our content was then filtered through an AI content generation engine, which then reconstituted it all and spit it back out using a bunch of digitally homogenized words slathered over an already soggy sales pitch.
(As an aside: F- you AI; let’s see you come up with a phrase like “…digitally homogenized words slathered over an already soggy sales pitch”).
Truth be told, it wasn’t just our stuff they had scraped. It was a ton of companies. Jax Bourbon Social in Jacksonville, Creative Lending in Albuquerque, OK Tires in Shorncliffe, just to name a few. Honestly, I have no clue who any of these folks are but, if I’m to believe the article about me on that website, they’re just some of my clients at my new job. (By the way, I’m looking forward to meeting you all and doing some cool human-made stuff together).
The most disturbing part about this website, however, was not the scraped content. It was the timing. According to the Wayback Machine Internet Archives, the site was created in April, the same week as Dark Ages 2.0 published. In less than a month it already has years’ worth of content on it. With new stuff being posted multiple times a day.
And so it begins…
When I wrote Dark Ages 2.0 I hadn’t taken into consideration that today’s AI can’t really generate anything original. It relies on information that’s already in existence. It also doesn’t have any morals, so it doesn’t recognize stealing, impersonating, or lying.
Honestly, if I were going to start a new agency today, I’d want that agency to look a lot like The M Network. I’d love to have our clients. I’d love to have our footprint and reputation. And if I had no morals, I’d probably just let the machines steal all The M Network’s stuff and present it as if it were mine. Why do the hard work when AI can steal it for you???
I’m calling this phenomenon fAIking. I figure that if The M Network’s content is going to be scraped, repackaged, and used as a sales tool for some BS website, I should at least get naming rights to the scam.
FAIking is the act of using AI to impersonate others, steal their content, and present it as your own. When it happens to you, you’ve been fAIked. This is the online content equivalent of identity theft. It’s also fraud and copywrite infringement (we’ll be pursuing both – but that will likely be a fool’s errand).
SO… what can you do about this?
The short answer is nothing. None of the security measures installed on our website stopped us from getting scraped.
Based on the information our IT team has acquired, there are a number of different things called robot.text directives that will disallow certain bots access to your site – but not all of them. They are also considering doing something called “rate limiting.” This apparently puts a cap on the number of inquiries a source can make on your website – so an AI bot won’t be able to scrape the entire site… just some of it.
Will any of that work?
I have no idea. We’ll let you know.
In the meantime, you should be regularly monitoring the internet for mentions of you and your organization and have the tools in place to catch if fAIking happens so you can alert others. When we found out, we also filed a fraud report with the Federal Trade Commission and are awaiting to hear back (we’ll update this article should anything change).
We tried to find out who owned the fAIk site but all that was hidden. We talked to our corporate attorneys, but they can’t really serve cease and desist to people they can’t find or contact a site that’s run by bots. The big lesson, I suppose, is that you can do a lot of things that cost time and money, but the most likely and most uncomfortable thing you should do is get used to it. At the dawn of AI, fAIking is likely to get a lot worse before it gets better.
Good luck and Caveat Emptor
Thom Mozloom
Real Human and CEO of only one company
The M Network.
