AI Detection

How AI detectors work and how to navigate them: detection signals, false positive risks, and what the scores mean for academic writers.

35 articles

AllAI DetectionAI HumanizationAcademic WritingSEO
Compilatio AI Detector Review: Europe's Institutional Detector
AI Detection

Compilatio AI Detector Review: Europe's Institutional Detector

Compilatio is the detector European universities license where American roundups assume Turnitin: French-built, used by more than 1,100 schools across over 50 countries, and the only major institutional detector a student can run on their own work before submitting. It claims a 94 to 99% reliability rate at catching AI text and under 1% false positives. The one peer-reviewed study to include it ranked it second out of 14 tools, and concluded the tools it tested were neither accurate nor reliable. Here is what the evidence supports.

Moe
8 min read
Sapling AI Detector Review: A Developer Tool in an Academic World
AI Detection

Sapling AI Detector Review: A Developer Tool in an Academic World

Sapling's AI detector comes from an NLP tooling company, not an integrity company, and it shows: a public API, per-sentence perplexity scores and a Chrome extension built for triage work. Its independent record is the widest we have reviewed, from zero false positives in Scribbr's test to a 90 percent false positive rate in a 2025 Texas A&M study. That volatility, not any single score, is the finding.

Moe
7 min read
Winston AI Review: The 99.98% Accuracy Claim vs the Evidence
AI Detection

Winston AI Review: The 99.98% Accuracy Claim vs the Evidence

Winston AI advertises 99.98% accuracy, the highest self-claimed figure in the detector industry, measured on an internal test set the company has never published. The peer-reviewed RAID benchmark put its overall accuracy at 71% at a fixed 5% false positive rate, still second among the four commercial detectors tested. Here is what each number actually measures, and who this tool genuinely suits.

Moe
7 min read
Scribbr AI Detector Review: What a 78% Score Actually Means
AI Detection

Scribbr AI Detector Review: What a 78% Score Actually Means

Scribbr's free AI checker posts the best free-tier numbers around: 78 percent accuracy, tied for first, in Scribbr's own published comparison, with the premium tier at 84 percent. The catch is the word "own": the test was designed, run, and graded by the vendor, and the other top free tool belongs to the same parent company. Here is what those numbers rest on, what the premium tier adds, and why a Scribbr pre-check cannot forecast a Turnitin result.

Moe
7 min read
Copyleaks AI Detector Review: The Detector Built Into Your LMS
AI Detection

Copyleaks AI Detector Review: The Detector Built Into Your LMS

Copyleaks is the AI detector most students meet without choosing it, bundled with plagiarism checking inside Canvas, Moodle, and Blackboard. The vendor advertises over 99 percent accuracy and a .03 percent false positive rate. In the June 2026 Vrije Universiteit Brussel study, it failed to fully flag a single one of 40 entirely AI-generated papers. Here is what the evidence actually shows.

Moe
6 min read
Pangram AI Detector Review: The Tool Researchers Keep Validating
AI Detection

Pangram AI Detector Review: The Tool Researchers Keep Validating

Pangram publishes a false positive rate of roughly 1 in 10,000, and unlike most detector vendors it now has recent independent research pointing in the same direction. A peer-reviewed 2026 study from Vrije Universiteit Brussel found it was the only tool of four that produced satisfactory results on master's level papers. Here is what the studies measured, where the evidence is still thin, and what no detector score can prove.

Moe
7 min read
Originality AI vs Turnitin vs Copyleaks compared on buyer, published accuracy claim and what each report shows
AI Detection

Originality.ai vs Turnitin vs Copyleaks

Three detectors, three buyers, three very different standards of published evidence. Turnitin arrives through a university's admin settings, Originality.ai is bought with a card on a credit plan, and Copyleaks reaches students through LMS integrations. What each one measures, and which one is likely to score your work.

Mark
7 min read
AI detectors compared by who each one is sold to and whether a named study sits behind its accuracy figure
AI Detection

The Best AI Detectors for Academic Writing in 2026: 10 Tools Compared

Two of these ten detectors publish the same 99.98% headline accuracy figure. One names the researchers behind it and breaks its false positive rate out by genre. The other names no test at all. Ten tools placed on four axes: who buys them, what they claim, what evidence sits behind the claim, and what the report actually shows, plus the process-tracking model that may replace scoring altogether.

Mark
15 min read
Table comparing GPTZero vs Turnitin on who can run each one, what a marker sees, and what each one costs
AI Detection

GPTZero vs Turnitin: Which One Decides Your Grade

GPTZero and Turnitin get treated like two versions of the same test. They are not. One is licensed to your university and attached to your actual submission. The other is a consumer product you can run yourself at midnight. Here is what separates them.

Mark
6 min read
Diagram contrasting a hidden-character watermark with a sampling watermark carried in the model's word choices
AI Detection

The Claude Watermark: Where the Mark Lives and What It Proves

Anthropic began marking Claude's text output on 11 August 2026. Almost every explanation of it since has described the wrong mechanism. The mark is not a character hidden in the page, it is a statistical bias in which words the model picked, and that distinction decides everything: what carries it, what clears it, and how little a positive result actually settles about who wrote a document.

Mark
10 min read
ZeroGPT accuracy: the company's own claimed rate next to its disclosed limitations on formulaic writing
AI Detection

Is ZeroGPT Accurate?

ZeroGPT's own FAQ claims a rate pushing toward 98 percent on internal testing, and separately admits that formulaic writing can read as AI regardless of who wrote it. Here is what the company states about its method, what independent testing on this exact product has found, and what a free score is actually useful for.

Moe
4 min read
Illustration for an academic integrity AI guide, showing a student's course handbook next to a chatbot conversation window
AI Detection

Academic Integrity and AI: A Student's Guide

A practical walk through what AI use is almost always fine, what is almost never fine, the grey area in between, and how to keep a simple record of what you did, starting from the one policy that actually governs an assignment: the course's own.

Moe
5 min read
Turnitin false positive rate: table comparing document-level and sentence-level figures and the sub-20-percent asterisk threshold
AI Detection

Turnitin False Positive Rate: What the Numbers Say

Turnitin publishes two false positive figures, not one, and neither describes the paper sitting in front of one particular professor. Here is the document-level versus sentence-level split, the sub-20-percent asterisk, and the arithmetic that turns a fraction of a percent into a real count of students.

Moe
5 min read
Illustration explaining AI detection in nursing school, showing a reflective writing assignment beside a fitness to practise case file
AI Detection

AI Detection in Nursing and Medical Programs

Nursing, medicine and allied health programs run academic misconduct and fitness-to-practise as two separate systems, so an integrity finding does not always stop at a grade. Here is how that two-track structure works, and why reflective writing and care plans are exactly where it gets tested.

Moe
4 min read
Universities that stopped using Turnitin's AI detector: Vanderbilt, Michigan State and UT Austin opt-out timeline
AI Detection

Which Universities Stopped Using Turnitin's AI Detector

Vanderbilt, Michigan State and UT Austin have each published their own reasoning for switching off Turnitin's AI writing indicator. None of them dropped Turnitin itself. Here is the distinction, the arithmetic behind it, and which other names get repeated without an institution's own page to confirm them.

Moe
5 min read
Illustration answering do college admissions check for AI, showing a personal statement next to a signed application declaration
AI Detection

Do College Admissions Check for AI?

Common App calls AI-generated essay content fraud, UCAS requires a declaration and runs its own similarity checks, and Brown allows AI for spelling and grammar only. Here is what admissions offices have actually published, and why a personal statement is judged on voice long before any scanner gets involved.

Moe
5 min read
How to prove you didn't use AI: a table of evidence types from version history to search logs and how hard each is to fake
AI Detection

How to Prove You Did Not Use AI

There is no single fact that proves how an essay was written. What exists instead is corroboration: independent, hard-to-fake traces that converge on the same account. Here is what that evidence actually looks like, and what to start doing now if you have not been accused of anything yet.

Moe
6 min read
Table answering can Turnitin detect DeepSeek, Claude and Gemini, based on Turnitin's own published model coverage
AI Detection

Can Turnitin Detect Claude, Gemini and DeepSeek?

A classifier trained to react to statistical pattern does not need to have seen a specific chatbot to flag its output. Here is why that matters for DeepSeek, what Turnitin itself publishes about Claude, Gemini and other models by name, and what it does not publish for any of them.

Moe
4 min read
Accused of using AI on essay work: a calm first 48 hours checklist for responding to an academic integrity email
AI Detection

Accused of Using AI When You Did Not: What to Do

An academic integrity email naming AI use is frightening precisely because there is no single fact that disproves a percentage. Here is what to do, in order, in the first 48 hours: what to say, what never to admit, what evidence to request, and how to find out whether a score is even allowed to stand alone.

Moe
6 min read
Diagram answering can Turnitin detect paraphrased text, comparing similarity matching against the AI writing indicator
AI Detection

Can Turnitin Detect Paraphrased Text?

Yes, Turnitin detects paraphrased text, and it was built to. Its similarity engine catches reworded passages that still overlap with published and previously submitted work, and its AI writing report carries a labelled category for AI-generated text revised with a paraphrasing tool or word spinner. Here is how each system catches it, and why swapping words for synonyms clears neither one.

Moe
9 min read
How does Turnitin detect AI: diagram of the segment, score and average pipeline behind a Turnitin AI writing report
AI Detection

How Turnitin Detects AI, Step by Step

Turnitin's AI writing detector does not read a document as a whole. It segments a submission into chunks, scores each one, averages the results, and only then produces the percentage on a report. This covers that pipeline, what it flags across ChatGPT, Claude and Gemini, what independent testing found it misses, and who actually gets to see the number.

Moe
18 min read
What is a good Turnitin score? A similarity report broken into matched sources rather than one number.
AI Detection

What Is a Good Turnitin Score?

Turnitin publishes no acceptable similarity percentage, and institutions set their own guidance instead. This post explains what actually drives a high or low score, why quoted material and reference lists inflate it without anything being wrong, and how to read the match breakdown behind the number rather than the number itself.

Moe
6 min read
Before and after example showing how to increase burstiness and perplexity by varying sentence length and word choice
AI Detection

How to Raise Perplexity and Burstiness in Your Writing

Raising perplexity and burstiness on purpose usually just means writing better: varying sentence length, mixing clause structures, and choosing the word that actually fits instead of the one that merely scans. Concrete before-and-after examples, plus where this stops being craft and starts being a bad idea.

Moe
10 min read
Diagram showing how GPTZero works today: a sentence-level AI classifier replacing its original perplexity and burstiness method
AI Detection

How GPTZero Works

GPTZero made perplexity and burstiness famous, then quietly retired both. Here is what its classifier measures now, what a report actually shows, and where the tool says its own limits are.

Moe
5 min read
Diagram illustrating token probability ai detection: a language model's next-token distribution turning into a log-likelihood score
AI Detection

Token Probability and Log-Likelihood in AI Detection

Perplexity gets the attention, but the arithmetic underneath it is token probability: what the model's distribution over the next word actually contains, why detection math runs in log space instead of raw probability, and how a string of per-token scores becomes one number.

Moe
6 min read
Is burstiness still used by AI detectors? GPTZero's 2023 shift from a statistical formula to a trained classifier.
AI Detection

Do AI Detectors Still Rely on Perplexity and Burstiness?

GPTZero made perplexity and burstiness famous, then quietly retired both in autumn 2023 for a deep-learning classifier. This post traces what changed, what GPTZero and Turnitin actually document today, and why the two statistics still explain why machine text is detectable even though neither vendor computes them directly anymore.

Moe
5 min read
Illustration answering will I get caught using ChatGPT, showing the stages of a university academic integrity process from report to appeal
AI Detection

Will I Get Caught Using ChatGPT?

Getting caught is rarely a single moment triggered by a percentage score. It is a defined process with stages, evidence standards and appeal rights, most of which never make it past a first conversation. Here is what that process actually looks like end to end, using a published university procedure rather than a guess.

Moe
8 min read
Illustration answering can professors tell if you use ChatGPT, showing a professor comparing a new essay against a student's earlier drafts
AI Detection

Can Professors Tell If You Used ChatGPT?

Detection software matters less here than most students assume. This walks through what a professor who has already read your work actually notices when a draft stops sounding like you, from a landmark blind study of examiners to the specific, checkable signs that show up long before anyone runs a scan.

Moe
6 min read
What is burstiness? A chart comparing uniform AI-generated sentence lengths to varied human sentence lengths.
AI Detection

What Is Burstiness? Why Uniform Sentences Get Flagged

Burstiness measures how widely sentence lengths swing across a passage, not how long the average sentence runs. This post explains where the term comes from, how the spread is actually calculated, and why a language model's output tends to settle into a narrow band of lengths.

Moe
6 min read
What is perplexity in AI detection? A language model scoring how predictable each word in a sentence is.
AI Detection

What Is Perplexity in AI Detection?

Perplexity is the number a language model produces to describe how surprised it was by your word choices. This post traces the metric to its 1977 origin, works through the formula, and explains why the same passage can score differently depending on which model is doing the reading.

Moe
6 min read