Flagged as 97% AI
The algorithm doesn't lie. But it does charge $49.99 a month.
Brandon Choi’s thesis was rejected twice. One of those times, it was for being AI-generated.
Three years of his life, gone. Researching, writing, and revising “Authentic Voices: Identity and Narrative Truth in Contemporary Immigrant Literature.” One hundred and fifty pages analyzing how second-generation Korean American authors negotiated the tension between cultural heritage and assimilation. Brandon read every major text in the field. He conducted original interviews. Seventeen full revisions.
The email from the Crestview State University MFA committee arrived on a Tuesday morning.
Subject: Thesis Submission Update - ACTION REQUIRED
Dear Mr. Choi,
Your thesis submission has been flagged by our institutional plagiarism detection system, SubmitIt.ai, with an 89% probability of AI generation. Per university policy adopted in Fall 2023, all graduate submissions must score below 20% to be considered for approval.
We take academic integrity seriously. Please schedule a meeting with the committee to address these concerns.
Brandon read it three times.
No AI. Not even Grammarly. Every word belonged to him.
The committee meeting was worse than he expected.
“Mr. Choi,” Dr. Patterson said. She kept her eyes on her laptop. “The score is quite clear. Eighty-nine percent.”
“I wrote every word myself.” Brandon pressed his palms flat against the table. “I can show you my drafts. My research notes. I have timestamps.”
“The algorithm doesn’t lie,” Dr. Martinez said. “It’s trained on millions of texts.”
“But I didn’t...”
“We’re not accusing you of anything yet,” Patterson said. “We’re giving you the opportunity to revise and resubmit. Make it more... authentic.”
Brandon went home and read his thesis again. Maybe it was too polished. Too formal.
For three days, he butchered his own work. He dumbed it down. Contractions replaced formal phrasing. Sentence lengths varied aggressively. Awkward personal asides about his own identity cluttered the pages. He removed his most elegant transitions and replaced them with blunt force. It felt like taking a key to the paint job of a new car, but he needed to graduate.
He resubmitted with a note: Please re-test. I’ve revised to address any concerns about authenticity.
The committee ran it through SubmitIt.ai again.
Result: 93% AI-GENERATED.
The rejection email arrived the next day:
Dear Mr. Choi,
We must delve into the concerns surrounding your submission. The tapestry of your argument, while navigating the landscape of contemporary literary discourse, raises questions about the authenticity of your voice. We leverage our expertise to ensure the integrity of our program remains robust...
We are disappointed that the revisions have not addressed our concerns. In fact, the increased score suggests you may be using increasingly sophisticated AI tools to evade detection.
Your appeal is denied.
Brandon stared at the email. The words “delve” and “tapestry” glared at him.
He copied the email, opened a browser, and pasted it into the SubmitIt.ai public demo.
Result: 97% AI-GENERATED.
He replied immediately: “This rejection letter appears to have been written by AI.”
The response came within an hour:
Mr. Choi, we are asking the questions here, not you. Please refrain from further communications on this matter.
Sleep was impossible. At 2 AM, Brandon sat at his laptop, researching SubmitIt.ai.
Then he found HumanWrite Pro.
Same company. Same website footer. Same corporate address in Delaware.
“Did SubmitIt.ai flag your work? HumanWrite Pro can help! Our advanced algorithms rewrite your text to pass AI detection. Guaranteed results.”
Pricing:
Basic: $19.99/month (10,000 words)
Academic: $49.99/month (50,000 words, Guaranteed <5% SubmitIt score)
Professional: $99.99/month (Unlimited, Enterprise-grade humanization)
The detector company sold the solution to the problem it created.
Brandon dug deeper.
SubmitIt.ai – founded 2023, Veridian Trust Technologies
HumanWrite Pro – founded 2023, Veridian Trust Technologies
TextVerify Labs – founded 2023, Veridian Trust Technologies
One company owned them all. The detectors and the “fixers.”
Then he found the venture capital connection. Veridian Trust’s primary investor: Cascade Ventures.
Cascade Ventures’ portfolio:
ClosedAI (makers of ShitGPT)
Anthropomorphic (makers of Clod)
Y (makers of Brok)
The same investors who funded the AI that “threatened” academic integrity were funding the detectors that “protected” against it. And the services that “fixed” it.
It wasn’t a bug. It was the business model.
“Nice thesis you got there. Shame if someone thought it was a bot. For $49.99 a month, we’ll make sure that doesn’t happen.”
Higher education had always been pay-to-pass. Now it was just more honest about it.
Brandon spent the next week writing two thousand words. Carefully researched. Every claim documented. Screenshots of the corporate connections and analysis of the financial incentives filled the pages.
He titled it: “The AI Detection Scam: How One Company Profits From Both Sides.”
Submission to Minimum resulted in a rejection within six hours: “This submission has been flagged as AI-generated spam content.”
He revised it, added his personal story, and pitched it to the Chronicle of Higher Education. The reporter responded: “This is a compelling story, but we cannot publish an article that itself fails AI detection. The irony isn’t lost on us, but editorial policy is clear. Your piece scores 91% on SubmitIt.ai.”
Brandon tried the university newspaper. The student editor was sympathetic.
“Dude, I believe you,” he said over coffee. “But the university installed SubmitIt.ai campus-wide last semester. I literally can’t publish anything that gets flagged.”
A quick scan of Brandon’s article through the system confirmed it. 94% AI-GENERATED. “See? I’m sorry, man.”
Brandon created an account on Y (formerly Tweeter). If traditional media wouldn’t publish him, he’d publish himself.
He wrote a 47-tweet thread detailing every connection. The thread went semi-viral, with 2,400 views and 38 retweets.
Three hours later, a Community Note appeared:
Reader Context added by Brok:
This user’s post has been flagged as high-probability synthetic media by Brok, Y’s AI assistant. Fun fact: Humans typically make 3.4 typos per 1,000 words! This thread contains zero typos across 2,847 words.
Engagement dropped to zero. The algorithm had spoken.
Fine. His own platform. Complete control.
Brandon created a blog titled “Authentic Voices.” He posted the full exposé and promoted it on Reddit, LinkedIn, and Facebook.
Two weeks later, he checked Google Search Console.
Status: Indexed but not ranking.
He searched manually: “AI detection scam”
Page 57.
“Veridian Trust Technologies”
Page 71.
Google’s search algorithm used Veridian Trust’s detection technology, giving low-quality “AI” content an SEO death sentence. His carefully researched, grammatically perfect exposé sat algorithmically buried.
The pattern was clear. Coherent arguments looked like AI. Well-structured prose looked like machines. Zero typos looked suspicious.
Good writing was indistinguishable from AI. Bad writing passed as human.
Brandon stared at HumanWrite Pro‘s homepage at 3 AM.
Academic Tier. $49.99/month. “Guaranteed <5% SubmitIt score.”
He knew it was the scam. The trap. But his exposé needed to be published. Someone needed to know.
Brandon entered his credit card information and uploaded his article.
“Making your writing authentically human... Processing...”
The progress bar filled slowly. When it finished, he opened the result.
Original:
“Veridian Trust Technologies has engineered a closed-loop financial ecosystem, effectively monetizing the proliferation of synthetic media while simultaneously controlling the market for its detection. This vertical integration creates a perverse incentive structure where the company profits directly from the degradation of academic integrity.”
Humanized Result:
“So, like, computers are crazy right? Money is also a thing we all use. This big company is doing business. It is like when you buy a hot dog but you also have to buy the bun from the same guy, and maybe the hot dog is fake? Business is complicated. We are looking at this today to see if trust is good or bad.”
It read like a recipe blog. Like those articles where you scroll through twenty paragraphs of filler before getting to the instructions on how to boil an egg.
Brandon stared at the screen for ten minutes.
Then he submitted it to Academic Freedom Quarterly, a tiny online journal that seemed to accept everything.
SubmitIt.ai score: 4% AI-detected.
Status: APPROVED FOR PUBLICATION.
It was published three days later. The hate-filled emails came quickly:
“Is this satire?”
“I genuinely cannot understand what are you trying to say.”
“Did ShitGPT write this ironically?”
His carefully researched exposé had become a meme. The protection racket remained unexposed. The only version that passed was the one nobody could understand.
The truth had become unpublishable by being well-written.
Brandon Choi had one option left.
His thesis. Three years of work and one hundred and fifty pages.
He uploaded it to HumanWrite Pro Academic.
$49.99.
“Processing your document... This may take several minutes...”
He waited twenty-three minutes.
The result appeared. His literary analysis of authentic immigrant voices, transformed into... something else.
He didn’t read it. He couldn’t. Brandon downloaded the file and resubmitted it to the committee.
SubmitIt.ai score: 6% AI-detected.
Status: APPROVED.
Three weeks later, the email arrived:
Dear Mr. Choi,
While your revised thesis now passes our authenticity verification requirements, the committee cannot approve work of such poor literary quality. The prose is incoherent, the arguments are unclear, and the narrative structure has completely collapsed.
We note significant degradation in quality from your original submission and must question whether you possess the writing ability expected at the MFA level.
We recommend substantial revision before resubmission.
Brandon copied the rejection letter and pasted it into SubmitIt.ai.
94% AI-GENERATED.
Brandon Choi withdrew from Crestview State University’s MFA program in Spring 2024.
He currently works as a content moderator for SubmitIt.ai, reviewing flagged submissions for $34,000 per year.
His thesis remains unfinished.
His exposé remains online, unread, on page 57.
If ‘Flagged as 97% AI’ made you uncomfortable, I have more UFiction: Sentenced to Heaven - 13 Episodes of Divine Judgement wrapped in bureaucratic satire. What if the judge goes to Hell for judging? What if you will find yourself in that piece, too?
THE END
This story is as haunting as a black mirror episode 😭 It has that same surreal is-this-fiction-or-is-it-the-future feeling. Great work as usual!!
I am a Masters student and my last assessment was flagged as “AI” (it wasn’t). The only way I could save myself was to show previous work of high-quality from pre-AI assessments. They agreed it was my own writing, but now I’m terrified of it happening again.