Imagine spending years developing your writing voice, only to reach a point where a machine tells you it sounds artificial. Worse still, you start to believe it.
This is the uncomfortable reality facing writers and creatives today. We have entered an era in which people question their own abilities because AI detectors suggest their original work looks machine-generated. Some have even started changing how they write, stripping away their natural style to satisfy an algorithm.
I've been writing since I was a seven years old, so this feels head-shaking. Writing has always been personal to me. It reflects how I think, how I see the world, and how I choose to express myself.
I have spent years developing that voice. I refuse to believe writers should now have to dilute it to prove their humanity.
And that is precisely the conversation I want to have.
When original writing becomes "suspicious"
The problem goes far beyond receiving a percentage next to a piece of writing. The more serious issue begins when writers start changing themselves to satisfy a detector.
Original work can receive an AI score because it appears clean, structured, articulate, or follows patterns that a particular system associates with machine-generated text. And those machine parameters keep changing!
A writer can spend hours developing an article from their own experiences and research, write every word themselves, and still receive a result suggesting that part of their work came from AI. For many people, that creates immediate doubt, even when they know exactly how they created the work.
Then comes the new, crazy, head-shaking editing process. Writers start removing phrases they would naturally use; they change sentence structures, avoid repetition, and deliberately make their work feel less polished. Some writers even disrupt their natural rhythm because they think a detector will interpret imperfection as more human.
That should concern every writer and creative professional, because we shouldn't have to distort original work to persuade a detection tool that a human created it. That's where the conversation around AI detection has taken us.
Shakespeare's work has reportedly received AI scores ranging from low percentages to dramatically higher ones depending on the tool used
Patterns have always been present in human writing. Creatives and writers form habits. We learn how to communicate more clearly, strengthen our ideas, refine our language, and return to structures that feel natural. These designs come from years of craft development; they do not turn people into machines.
This is also where we need to understand what an AI detector actually does. It does not look at a piece of writing and determine who created it. It makes a prediction based on patterns, assumptions, models, and parameters. A prediction is not proof, period.
Even MIT has highlighted concerns about the unpredictability and limitations of AI detection. Different tools use different approaches, parameters, and settings, so the same piece of writing can get very different results depending on which detector you choose.
We have seen this problem exposed in some of the most extreme examples. Shakespeare's work has reportedly received AI scores ranging from low percentages to dramatically higher ones depending on the tool used. Of course, Shakespeare did not use AI, but this inconsistency reveals a weakness in treating detector scores as definitive proof.
We have to think about what those percentages actually represent if one system finds a tiny amount of AI in a text while another finds a considerably greater amount in the same text. All too frequently, they are viewed as facts rather than estimates, and authors suffer the consequences.
The moment writers start losing their voice
I've mentioned in previous pieces that I have watched people I admired lose parts of their voices. I have seen writers with distinctive styles become increasingly cautious, stripping away the very characteristics that once made their work recognisable.
The process often starts with good intentions. A writer wants to avoid problems, questions, or accusations. They want their work to pass a detector. So they begin changing the language they naturally use and adjusting their writing around what they believe a machine expects from a human.
Because people have patterns, writing does too. Your thought process shapes your writing style, and your voice reflects your everyday routines.
Your personal experiences and learning shape your vocabulary, your favourite authors shape your rhythm, and your personality shapes what you choose to share with the world.
We have reached a point where people can create original work and still doubt themselves because a percentage on a screen tells them something different
No two writers bring the same combination of experiences to the page. Two people can write about the same event and tell entirely different stories because they notice different details and bring different experiences to the work.
That individuality gives writing much of its value. It also explains why the current culture of suspicion is a red flag for me. We have reached a point where people can create original work and still doubt themselves because a percentage on a screen tells them something different.
Think about what happens in that moment. Someone sits down with a blank page, draws on their own life, researches their ideas, constructs their own arguments, and writes every word themselves. Then a detector produces a result, and suddenly the writer starts questioning work they know came from them.
That percentage can hold more power than it deserves. The technology does not know your creative process. It does not know how long you spent developing an idea or how many drafts you wrote before reaching the final version.
A machine sees patterns, and makes a prediction.
What makes us unique as human beings often comes from what we can't standardise. Our voices carry habits, emotions, inconsistencies, experiences, and perspectives. Those qualities develop over a lifetime, and no detector can fully understand the person behind the words.
We still have human voices, let's not lose them
That doesn't mean we should ignore the challenges around AI-generated content. AI has changed the writing landscape, and people deserve honest conversations about how companies and individuals use these tools: transparency is important, and clear authorship standards matter.
We also need to protect something very fundamental. Writers should not have to make their writing worse to appear more human. They should not have to abandon their natural style because a prediction tool considers clarity, structure, or consistency suspicious.
As this technology continues to evolve, we cannot forget something important. We still have human voices. AI does not
We are doing something wrong if teaching people to strip themselves out of their own work is the answer to faulty detection. A person’s fear of what an algorithm might say about a paragraph should not erase years of developing their writing voice.
Remember, our voices are valuable because they come from us. They reflect how we see the world, what we have experienced, and the choices we make when we sit down to create something.
AI can generate language, but it does not live our lives. It does not carry our childhood memories, our private frustrations, our relationships, our failures, or the experiences that turned each of us into the remarkable human beings writing today.
Perhaps the most concerning consequence of this moment is not simply that AI can produce convincing writing. It is that some humans have started changing their writing to convince AI detectors that they are human.
We need to recognise how backward that has become. AI detectors can provide useful signals and form part of a wider review process, but they cannot replace human judgment, proof, context, or common sense.
They certainly should not become the reason a writer abandons a voice they have spent years developing. I nearly did too, but my common sense stopped me.
As this technology continues to evolve, we cannot forget something important. We still have human voices. AI does not.
So please, do not lose yours trying to prove yourself to a machine.
