Top-level FAQ
Frequently asked questions.
Quick answers, organized by topic. If yours isn't here, email hello@aiessaydetector.ai.
About the detector
What the tool does, how accurate it is, and where it fails.
How accurate is your AI detector?
On our internal benchmark (mixed native / ESL / AI-assisted English essays, 3,000+ labeled samples) we report 94.1% precision and 88.6% recall at the default threshold. Full methodology and broken-out numbers on /stats.
Does it work on short passages?
We return a confident score on passages of 250 words or more. Below that, we return a 'signal insufficient' result rather than a misleading number.
Which languages are supported?
English (US and UK), Spanish, French, German, and Portuguese. Non-English languages are scored using language-specific models, not a single multilingual classifier, accuracy varies by language and is published on /stats.
Is it biased against ESL writers?
Every commercial AI detector, including ours, has a measurable false-positive bias against non-native English writers. We publish the number honestly (currently about 2.3x the native-speaker false-positive rate) and continue to work on reducing it. See our methodology page for full disclosure.
Pricing & accounts
What's free, what isn't, and how our plans work.
Is there a free tier?
Yes. 5 checks per day, up to 1,000 words each, without an account. A free account raises limits to 20/day and adds sentence-level highlighting and CSV export.
Do you offer student pricing?
Students with a .edu email get 50% off any paid plan. Email hello@aiessaydetector.ai with your .edu address.
Do you offer classroom or institutional pricing?
Yes, teacher plans start at $12/month and include up to 150 student submissions. Institutional (school- or district-wide) pricing is custom; request on /for-institutions.
Can I cancel anytime?
Yes. Cancellations take effect at the end of the current billing period. No cancellation fee.
Privacy & data
What we keep, what we don't, and how data is protected.
Do you train on my essays?
No. We do not use customer-submitted text to train any detector or language model. Submitted text is retained for 30 days for abuse-detection and rate-limiting purposes, then deleted. Full detail on /privacy.
Do you share data with third parties?
No, with two narrow exceptions: (1) our infrastructure subprocessors (Cloudflare for hosting, listed in our DPA), and (2) law-enforcement requests with valid legal process, which we disclose in an annual transparency report.
Is the tool GDPR-compliant?
Yes. We have a DPA available at /dpa and honor all EU/UK data-subject rights including access, portability, and deletion requests.
Where is my data stored?
Primary storage is Cloudflare's EU region for EU users and US region for everyone else. No data is intentionally transferred across regions except for abuse-detection metadata.
Humanizer
What the humanizer is for, and what it isn't for.
Is using a humanizer cheating?
It depends on what you're humanizing. Editing your own AI-assisted draft to match your voice is normal editing. Running a 100% AI-generated essay through a humanizer to pass a detector is cheating under almost every academic-integrity policy. The tool doesn't know which situation you're in, you do.
Will humanized text pass every detector?
No. Humanizers that beat one detector often fail against another, because different detectors weight different statistical features. The only way to reliably pass every detector is to have written the essay yourself.
Why is the humanizer gated behind a free account?
To reduce abuse. An account lets us rate-limit per-user and investigate patterns of misuse. Accounts are free and take 30 seconds.
For educators
How the tool fits into a classroom workflow.
Do you integrate with Canvas, Blackboard, or Moodle?
Canvas integration is in beta as of April 2026. Blackboard and Moodle are on the roadmap. Email us for early access.
Can I run multiple essays at once?
Yes, teacher and institutional plans include batch upload (up to 50 essays per batch) with CSV export of results.
How should I explain detector output to a flagged student?
Lead with the conversation, not the number. See /blog/teaching-in-the-age-of-ai for a full script.
Still have a question?
Email hello@aiessaydetector.ai, we answer within one business day.
Open the detector →