Real People, Fake Comments, and Everything In Between

Who’s ready for a plot twist and another rabbit hole adventure? ME, ME! And I hope you are as well.

It turns out my Bot or Not? post from the end of September isn’t the end of the story. If anything, that post seems to have invited a whole new cast of questionable characters into my comment section. Gone are the days when spam was easily recognizable by its disconnected response, awkward and often garbled phrasing, and obvious sales pitches. OG spam of that nature hasn’t completely disappeared, but now it’s being joined by comments that are specific, relevant, and sometimes surprisingly convincing.

And the more I’ve investigated, the more I’ve realized that figuring out whether a comment was written by AI might be the least intriguing question. Get yourself a beverage and settle in for an interesting read.

Well, This Is Awkward

One of the comments I featured in the Bot or Not? post came from someone identifying themselves only as BioTrendy. It was lengthy, remarkably detailed, and attached to a commercial website. Naturally, I had questions and doubts, which I shared in some detail.

Apparently, BioTrendy had something to say about that.

Imagine my surprise when Martin, who identified himself as the person behind BioTrendy, showed up in my comments to set the record straight. According to him, he’d written the original comment himself, was a regular reader of my blog, and had simply used his business name rather than his own.

Wait. What?

The last thing I want is for a genuine reader to feel unwelcome simply because I’ve become suspicious because of what’s showing up in my comment section. I so appreciated Martin taking the time to explain himself, and I made sure to let him know that.

I’d spent all that time questioning whether the comment was genuine, and now here was someone telling me that not only was it written by a real person, but that person was a regular reader! It certainly gave me something to think about. Had I been too quick to question a comment simply because it was unusually detailed and attached to a commercial website?

Or maybe the problem wasn’t my inability to recognize a real comment. Maybe I’d become so accustomed to spotting suspicious ones that I was starting to question the genuine ones, too. And here’s something else I can’t help wondering. Martin said he’d never imagined that leaving a comment on someone’s blog would inspire an entire post. How surprised will he be to discover that he’s now appearing in the sequel?

Sorry, Martin. It seems that you’re becoming a recurring character around here.

Meanwhile, the questionable comments haven’t stopped coming, and the collection is quite impressive. All have been written in the last 30-odd days. Some are laughably obvious, others are surprisingly convincing, and a few have sent me down rabbit holes I never expected to explore.

So, shall we?

Not Quite Picture Perfect

I’ll start with a seemingly harmless comment left in one of my older posts. Nothing particularly remarkable about it. Just a friendly compliment, the sort of thing any blogger would be happy to receive.

What really caught my attention was that the comment had been left directly on an image rather than on the blog post containing it. I hadn’t even realized that visitors could comment on individual images in my media library.

Was the image that warranted this comment deserving of such flattery? I’ll let you judge for yourself.

I’d say no. So, was this a genuine reader who somehow landed on the image attachment page, or someone looking for another way to place comments on my website? I can’t say for certain, but given that the post had been written over two years ago, I find the latter to be the more likely explanation. This comment also prompted me to investigate how WordPress handles comments on media attachments.

The result? I’ve since disabled commenting on individual images. Who knew that was a thing? Anyway, now there’s one less doorway for questionable comments to find their way onto my blog.

Facebook Friends?

Next came a series of comments on my January post, Finally Getting Around To It: I Give Up. Different names, different comments, but something rather peculiar in common: every commenter had included the exact same Facebook profile URL.

Now, I suppose it’s possible that several people could share a Facebook account, but that seems an unlikely explanation here. More likely, someone was using different identities while directing visitors to the same destination.

Here’s something worth knowing about WordPress comments. Anyone can enter a name, email address, and website URL without verifying that any of that information actually belongs to them. And when a comment is approved, that website can become a clickable link attached to the commenter’s name.

So, were these several different readers who just happened to share a Facebook profile? Or one person (or bot) pretending to be several different people?

Whatever the explanation, Akismet (my spam-filtering plugin) wasn’t buying it. All the comments landed in my spam folder, where they stayed until their automatic deletion 15 days later.

Call Me, Maybe?

Another commenter, also using a Facebook identity (same as above, so already suss), appeared interested in getting in touch with me. Under different circumstances, I might have taken the question at face value.

Except there’s a small problem. My blog does indeed have a Contact page, and it’s not exactly hidden.

Now, I know not everyone takes the time to explore a website before leaving a comment. Still, asking about something that’s readily available struck me as a possible attempt to get a response rather than a genuine request for information.

When It’s The Little Things

This next one was a bit more puzzling. The comment itself is very specific. It references my rugby-shirt rabbit hole, style numbers, MSRP tags, manufacturing countries, authenticated resale, and my point about an AI-generated answer sounding convincing. That is not generic “nice post” filler. Someone or something clearly processed the content quite well.

Interestingly, Akismet sent it to Pending rather than Spam, leaving the decision about whether to approve it entirely up to me. What made it suspicious to me was the identity information (not shown). The commenter called themselves “this store” and linked to a .store URL that is different than the one the brand uses for its official site. When I looked into it further, I couldn’t find the .store domain among the brand’s authorized retailers. When a website link is part of the comment package, it must also be scrutinized.

There’s one other thing about the comment that triggered my spidey-senses, though. Did you catch it as well?

It was the use of the phrase “tidy story” at the end. The full sentence “The physical tags always beat the tidy story” struck me as the kind of polished, slightly unnatural phrasing I often associate with AI. It’s entirely possible that a real person actually wrote that, but it sounds more crafted than conversational, especially in consideration of the rest of the comment.

Also, the em dash.

Irfan left another comment on a subsequent post that followed the same pattern: slightly unnatural phrasing, use of an em dash, and an even more significant plug for his website. Neither comment was approved.

Along a similar vein, these next two comments from someone identifying themselves as Asher Gray of the BagOrbit team seemed almost genuine. Both referenced specific details from my outfits, and neither sounded like the usual generic spam.

But then came the sales pitch. Both comments managed to slip in a BagOrbit product link into what otherwise appeared to be a thoughtful response to my post. Coincidence? I think not.

For a moment, I considered approving the comments after removing the links. After all, they were relevant to the post, and who doesn’t appreciate a little engagement? But that raised another question. If the comments were written primarily to promote a website, would removing the links make them genuine? Or would I simply be publishing advertising disguised as conversation?

A Familiar Name, An Unfamiliar Website

Just when I thought I had enough meat for this post, another example came my way. I’d like you to meet Tahir. Or at least that’s the name attached to two comments that landed in my spam folder two weeks ago, six hours apart, on the same post.

Both comments referenced specific outfits, but what interested me was how differently they were written. The first (lower one in the image) was relatively casual, written mostly in lowercase. The second (seen at the top) was more polished and included even more observations about my clothes.

My first thought was that Tahir might have returned to the post to see whether his original comment had been approved. When it hadn’t appeared, perhaps he decided to try again. Of course, that’s only speculation. What really caught my attention, though, was the website attached to his name: Croft & Barrow.

As a frequent Kohl’s shopper, I immediately recognized Croft & Barrow as one of its clothing brands. What I didn’t recognize was the website claiming to represent it. Curiosity got the better of me, so naturally I went looking.

The website appeared legitimate at first glance. There were clothing categories, product photographs, an About Us page, social media links, and even articles explaining what supposedly had happened to Croft & Barrow at Kohl’s. But the more I explored, the more questions I had.

The About Us page offered plenty of reassuring language about quality, comfort, and timeless style, but very little about the company actually operating the website. The linked social media accounts had virtually no followers or activity. Several articles, apparently published in January and February, discussed the brand and showcased its clothing, but I couldn’t find anything more recent.

And then there were the product links. Every listing I checked included a See Amazon option. Every one I tried led to a page-not-found message. So what exactly was this website selling? And why was someone using this website as their identity when commenting on my blog?

In my sleuthing, I also discovered an affiliate disclosure, suggesting that the website might be intended to generate commissions rather than sell merchandise directly. There’s nothing inherently wrong with affiliate marketing, but it didn’t explain the website’s apparent claim to be an official source for a brand I knew belonged to Kohl’s. By this point, I’d spent considerably more time investigating Tahir’s website than I had spent reading his comments.

And here’s a final bit of irony. During my investigation, I learned that Tahir is a name of Arabic origin meaning pure. Whether the intentions behind those comments were indeed pure is something I can’t answer. What I can say is that these two seemingly thoughtful comments led me to a website that left me with more questions than answers.

Who Am I Kidding?

Of course, questionable identities aren’t limited to my comment section. Sometimes I wonder about my own, particularly when I look at what I buy and what I wear.

Case in point: a floaty floral tulle skirt, a cable-knit cardigan, and western-style boots. Am I dressing like a teenager? Maybe. Do I care? Not particularly.

With my silvery hair, I’m certainly not fooling anyone into thinking I’m decades younger than I truly am. Nor is that my intention. But if there’s one fashion rule I’ve never been particularly good at following, it’s the one about dressing your age.

I actually bought this tulle skirt with the idea of removing the lining so I could wear the sheer layers underneath dresses or other skirts. But after putting this outfit together, I’m second-guessing taking it apart. For now, I’m liking it just the way it is.

The skirt follows through on the florals in my November Moodboard, while the olive henley carries over a color from October’s inspiration. The cream cardigan survived my recent Reduce by 50% sweater challenge, and the boots are part of my ongoing effort to wear more of the footwear already in my closet.

So maybe I’m not dressing my age. I’m just dressing according to what I like, which seems like a perfectly reasonable thing to do.

And now for the rest of this “plot twist and rabbit hole adventure” story.

Have I Poked The Bear?

Some of the most wack-a-doodle comments have been appearing on the very post where I first questioned whether my commenters were real people. It’s almost as though I’ve invited the suspects to participate in the investigation.

First up is one from “Daniel”, who offered his own advice on the topic. There’s something amusingly ironic about a suspicious commenter explaining how to recognize AI-generated comments.

Daniel made a surprisingly thoughtful observation: “Polished writing is easy to fake, a specific and slightly off-center reaction is much harder.” But the comment itself was unusually polished and analytical, almost as though AI had been asked to generate a convincing response to my post.

The bigger red flag was the commenter’s identification. His website, freevocalremover.ai, promotes an unrelated AI audio service, and his email address uses the domain @agent.qq.com. That made me wonder whether the comment was less about engaging in conversation and more about getting a commercial backlink. Since WordPress can make the commenter’s name clickable, there’s no need to include a promotional link in the comment itself.

Last up is this comment from someone identifying themselves as the GetVocalRemover team. Another website offering AI-powered audio tools? What are the odds? Like Daniel, this commenter responded with remarkable specificity, referencing details from my packing experiment and even asking what I’d ultimately done with one of the suspicious comments I’d discussed.

But this time, the most interesting clue wasn’t the wording or even the website. It was the email address domain. A little internet sleuthing revealed that mail.grokbot.com is a domain that provides email addresses to AI agents. Yes, AI agents can have their own email addresses.

As I mentioned earlier, anyone can enter an email address in a WordPress comment form without verifying that it belongs to them, so this doesn’t prove an AI agent wrote or submitted the comment. But an AI-related email address attached to a suspiciously knowledgeable comment on a post questioning AI-generated comments?

I’m truly beginning to think I may have indeed poked the proverbial bear.

Who’s Fooling Whom?

As with the initial blog post, the deeper I’ve dug, the harder it’s become to separate genuine engagement from something manufactured. Are these real people using AI to write comments? Are they automated accounts? Or are they simply very determined marketers hoping to get a little free advertising?

In case you’re wondering whether the old-fashioned spam has disappeared, of course it hasn’t. At the same time, though, the newer variety is becoming increasingly more and more difficult to recognize. Take a look at these two comments on my most recently published post, both of which landed in my spam folder. One hardly makes an effort to disguise its intentions, while the other could almost pass for a thoughtful response from a regular reader.

Same spam folder. Same apparent objective: a backlink. But one comment was so convincing that it took the mistakes to give it away. Can you spot them?

On second thought, perhaps I shouldn’t be so quick to point fingers. When I leave comments on other blogs, I typically include my own website URL. Doesn’t that make me guilty of trying to get a backlink as well? I’d like to think the difference is that I’m genuinely interested in what the blogger has written, but I suppose someone looking at my comment from the other side might have their doubts, too.

And here’s something else to consider. As the administrator of my blog, I certainly have the ability to edit comments before approving them. I could remove a suspicious website link. I could even replace an entire comment with something like, “Love, love, love! Such a fabulous post. Thanks for sharing!”

Of course, at that point, I’d be the one creating a fake comment. How’s that for a plot twist? 😂

Have you ever questioned whether a comment or online interaction was genuine? What were the clues that tipped you off?

Stay well and let’s keep in touch! Subscribe, follow, friend, pin … all options are available. I look forward to hearing from you soon.

~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~

Photo credits: MMPerez and me

Note: This is an unsponsored post; however, some or all of the links to clothing items may be associated with an affiliate program, where I earn a small commission for each purchase at no additional cost to you.