Imagine you spend hours writing a piece. You polish every sentence, fact-check every claim, and pour your voice into it. You hit publish, proud of your work. Then, within minutes, an algorithm flags it as AI-generated. You have no proof of innocence. No appeals process. Your reputation is shredded before you can even type a response.
This isn’t a hypothetical. It’s happening right now on Substack, where a new AI detection tool is turning writers into targets. The real problem isn’t AI-generated content flooding the internet. It’s that anyone can now be accused of using AI, and the accusation itself is the punishment.
I’ve been following the fallout on Substack, and the stories are both absurd and terrifying. One user posted a Note that was 50% AI and the detector said it was 100% Human. Another wrote a deeply personal essay about grief—only to see it flagged as synthetic. The tool doesn’t work. Nobody expects it to work. But it’s being used as a judge, jury, and executioner.
You’ve probably seen this wave of ‘AI detection’ services spreading across platforms. They promise to protect authenticity. They claim to separate human creativity from machine output. But here’s what they actually do: they weaponize suspicion against every creator, and they shift the burden of proof onto the innocent.
Let’s be clear about what’s happening. This isn’t about catching cheaters. It’s about creating a new power structure where platforms can accuse you with zero evidence, and you have no way to fight back. The tool is designed to fail—it’s a theater of certainty around an inherently unreliable technology. And the people who suffer aren’t sophisticated AI users. They’re honest writers, journalists, and artists who get caught in the crossfire.
I’ve seen this before. It’s the same logic that drove algorithmic content moderation on social media: false positives are acceptable collateral damage. But here, the stakes are higher. When a detection tool flags your work as AI, it doesn’t just hide your post—it questions your entire identity as a creator. And once that suspicion is seeded, it’s nearly impossible to uproot.
One Substack writer told me, ‘I felt like I had to prove I’m human. But how do you prove that? You can’t screenshot your own consciousness.’ That’s the trap. There’s no reliable appeals process because the detection itself is a black box. The algorithm is the authority, and the accused are left begging for mercy from a system that doesn’t understand human nuance.
So here’s the uncomfortable truth: AI detection isn’t a solution to a problem. It’s a tool for silencing voices under the guise of protecting authenticity. The platforms that deploy it are not interested in fairness—they’re interested in control. They want to decide who gets to be a ‘real’ creator and who gets flagged as a fraud.
We need to stop pretending this technology is reliable. It’s not. And the more we rely on it, the more we empower a new kind of gatekeeping that has nothing to do with quality or originality. The next time you see an AI detection score on your work, ask yourself: who is this tool really serving? Because it’s not you. It’s the people who want to keep you guessing, keep you afraid, keep you silent.
FAQ
Q: But doesn't AI detection help catch cheaters who use AI to mass-produce content?
A: In theory, yes. In practice, the false positive rate is so high that it punishes far more innocent creators than it catches actual cheaters. The tools are not accurate enough to be used as evidence, yet they are deployed as definitive verdicts. The 'cheaters' can easily adapt; the honest writers get trapped.
Q: What should I do if I'm falsely flagged by an AI detector on Substack or elsewhere?
A: First, document everything. Screenshot the detection result and your original drafts. Then reach out to the platform's support—but don't expect a quick fix. Most platforms lack a human review process for these flags. Your best defense is to build a visible track record of human work, but even that may not shield you. The system is fundamentally broken.
Q: Isn't the real solution just to accept that AI is part of the creative process and stop policing it?
A: That's a fair contrarian position. But the problem isn't about accepting AI—it's about the power imbalance. A detection tool that carries the weight of 'truth' without any accountability is dangerous regardless of whether you use AI. The fight should be about transparency and due process, not about banning or embracing specific tools.