The AI Witch Hunt: Accused of Using ChatGPT, Even When You Didn’t
Accused of using AI to write by the content police? You’re not alone.
“This was definitely written by AI.”
“This is AI slop.”
If you’ve written anything online lately, you’ve probably heard comments like these.
It’s a strange paradox. Tools like ChatGPT or Gemini can help non-native speakers or solo builders polish their ideas. It does make the process faster, especially helps with structuring what you want to say, and many legit writers use it in one way or another.
However, as the entry barrier for sharing your thoughts is low nowadays, tons of crap content is published everywhere: Medium, Reddit, even Substack.
And that hurts people who genuinely care about what they write.
If you admit using AI, you're almost labeled a fraud. No matter that the AI is just a small part of it.
If you don’t admit it? People start accusing you anyway. They will “notice” signs that it was written by AI. Using em dashes? You are done. Use cliche phrases? You are not a boring person anymore, you are an AI.
Welcome to the modern-day AI witch hunt.
Why People Hate AI Content
I get it.
I hate soulless AI-generated stuff too.
Especially when someone copies ChatGPT’s response into a blog post, adds a title, and hits publish. No editing. No thinking. Just a bunch of generalizations and SEO keywords.
What’s charming about the human written texts, is that all of them have unique style and ideas.
AI creates something that is just so boring to read.
Sure, it might seem more “perfect” than what an individual would write, but it’s also soulless.
Thanks to social media and fast dopamine, attention spans are shrinking. Fewer people read text at all. And AI is often used to create long posts, usually for SEO. That makes people even less likely to read, even when it’s good, human-written content.
Tough times for content creators.
I’ve Been Burned by This Too
When I say people “hate” this content, I mean some really do.
A year ago, I started a YouTube channel where I used AI voiceovers and some AI-generated images. I still put tremendous effort into editing. I enjoyed the process, but then came the comments:
It seems that people don’t care about the effort. If AI was used at all, then it’s automatically bad.
We’re All Sounding More Like AI
It’s funny that people are always trying to spot signs that something was written by AI: the repetitive phrases, generic intros, the weird punctuation.
But to be fair: AI writes this way because we do.
It just mimics the average stuff humans have written. So when people say, “that sounds AI-generated,” what they often mean is “that sounds like generic content.”
And the more we read AI content, the more we accidentally absorb its tone and structure. So we learn AI writing style, and start sounding like it too. Even though we’re not using it. That’s why figuring out if content is AI-generated, is becoming hard.
Nowadays, it’s very easy to get accused of using AI.
We’ve entered this weird phase where it feels like a modern-day witch hunt.
The Flawed Logic Behind AI Accusations
I already mentioned it in the intro, but here's the thing: AI often just improves grammar and structure. It lets non-native English speakers publish solid writing.
Back in the day, you’d find SEO content sites written purely to make money, and they were full of grammar mistakes. I remember one site listed on Flippa, supposedly helping people fight alcohol addiction, and it was so poorly written it was barely readable.
Now such writings are polished. So, overall, the same content that would exist either way, nowadays is more polished.
And when it comes to identifying this content, there are many articles on “how to spot AI writing”, and many of them are just wrong.
Take this paywalled post on Medium: This is How I Will Spot You Used ChatGPT.
The first red flag it gives is:
“In the modern digital age, content creation is more important than ever…
They marked this as definitely AI-generated.
But come on. That’s how a lot of people write, especially if they’re not experienced writers.
14-year-old me writing school essays would’ve started like that.
This just proves the point: AI is trained on us. It reflects our bad habits and clichés.
The Forbidden Hyphen and How to Spot AI Writing
Detecting AI-written content is hard. All these fancy AI checkers for ChatGPT, AI plagiarism detectors, AI detectors, whatever you call them, are not very effective.
A recent study confirmed that they can’t be trusted. Tons of false positives (human text detected as AI) and false negatives (AI passing as human).
Still, people look for signs like:
Em dash
A long dash (—) is a longer hyphen often used by AI to separate information in a sentence, expand on a clause, or indicate a break.
This is proper grammar, but many people do not even know how to find this symbol in their keyboards (on Mac’s its Option + Shift + hyphen (-). On Windows: Alt + 0151 ). It’s also a bit rare for a person to use this or hyphen overall, at least in casual blogging style.
Perfect grammar
No one’s perfect. Especially someone writing about the “best refrigerators” for an Amazon affiliate site. But AI? It smooths things out. It might rephrase for clarity, but sometimes sucks the life out of the writing. If it’s boring but grammatically flawless… it might be AI, or just a boring human.
Steady tone
People usually write in bursts. Some ideas are exaggerated. Some parts are more passionate. Real human writing shifts tone. And that’s what makes it interesting.
AI writing is often steady. But again, not always. A good human editor can make it steady too.
Factual mistakes
Let’s not forget AI still hallucinates. There’s nothing worse than a post positioned as written by an expert with factual mistakes.
There are many more, I will not go into details as some of them might be just wrong and too biased. It’s a tough task to evaluate if it’s an AI or human, when the AI was trained on humans work. There are plenty of AI detectors out there with people spending significant time improving them, yet they still fail.
Is everything written by AI nowadays?
Let’s not pretend.
Most content-heavy websites, especially those built for Amazon affiliate marketing, used to hire cheap writers from countries with lower rates.
This has not changed a bit. They still do outsource this, but now the AI is heavily used by an editor, which is then able to follow the guidelines and produce the content that is needed.
So even the big names and internet gurus often making millions from their content, do use AI. They might not personally, but their teams definitely.
It Does not Matter
If you provide value.
If you use AI to explain your ideas, polish your language and make a better syntax.
Personally, I don’t lack ideas. If anything, I have too much to say. I often write more than I need, then edit heavily or break it into two posts. And yes, I use AI for this part in order to make it more friendly for the reader.
How long does AI editing take? More than you would guess. It might take 10 prompts or more to make it solid.
I use Gemini or ChatGPT. I ignore most of what they suggest. Sometimes I ask for specific edits, then reject them too. Sometimes I keep the original.
But the key is: I’m still in control.
It doesn’t matter how it was made if your content delivers value, and if people get something out of it.