Bulk AI check
Review a class set for signs of AI writing in one upload
Upload the ZIP file downloaded from your learning platform to review submissions individually, open highlighted passages and export a CSV of the results. No plugin or administrative installation is required.
Illustrative results for fictitious students
- Amara Okaforessay-2.docxLikely written by AI91%
Flagged passages
The causes of the revolution were many, and historians still argue about them. In conclusion, it is important to note that the multifaceted nature of these events continues to shape modern discourse in profound ways. Bread prices rose in the summer of 1789.
- Liam Chenessay-2.pdfMixed writing signals46%
- Sofia Martinsessay-2.docxMixed writing signals22%
- Noah Fischeressay-2.odtLikely written by a person2%
- Hana Satoessay-2.docxLikely written by a person0%
Download CSV
How to review a class set for signs of AI writing
Fileverity’s batch checker analyzes submissions downloaded from Google Classroom, Moodle, Canvas, Microsoft Teams, Blackboard or Brightspace. Upload ZIP files or select individual documents. Each completed check returns an assessment (likely written by AI, mixed writing signals or likely written by a person) and the proportion of text flagged. Open a result to review the highlighted passages, or export the list as a CSV. Results are ordered by flagged proportion to help you plan your review. No plugin or administrative installation is required.
From downloaded submissions to individual results
- 01 / 03
Download the submissions
Download the assignment’s submissions from your learning platform as one ZIP file. The steps for each platform are below.
- 02 / 03
Drop the ZIP file
Open the batch checker and add the ZIP file or select individual documents. The ZIP is opened in your browser. Use Check only to filter folders and filenames by a term such as the assignment title.
- 03 / 03
Read the results
Each document is analyzed separately. Once the batch is complete, results are ordered by the proportion of text flagged. Open a row to review highlighted passages. The CSV export includes the filename, assessment, flagged proportion, likelihood, character count, credits and check ID.
Download a class set from your platform
Use the submission files your platform allows you to download. The batch checker runs separately from the platform and requires no plugin installation.
Google Classroom
Open the assignment in Classroom and use its Folder button to open the assignment’s folder in Google Drive, then choose More > Download. Drive prepares the folder as a ZIP file.

Moodle
Open Submissions for the assignment, then choose Download all submissions from the Grading actions menu above the grading table. Online text submissions are included as HTML files, which the batch checker can also read.
Canvas
Open the assignment page and choose Download Submissions in the right-hand sidebar. Canvas builds one ZIP file with each student’s latest submission.
Microsoft Teams
In the class team’s SharePoint site, open Student Work and download the relevant student folders. Add the ZIP file to the batch checker, then enter the assignment name in Check only to select the submissions you want to review.

Blackboard
Ultra: open the Gradebook, open the assignment, choose Download Submissions, select the students and choose Create ZIP file. Blackboard sends the link by email. Original: in the Grade Center, open the assignment column menu, choose Assignment File Download, select all and click Submit.

Brightspace
In Assignments, open the submissions, select all and choose Download. Brightspace prepares a ZIP file containing a folder for each submission.
Supported files and analysis limits
| Item | What applies |
|---|---|
| Files you can drop | ZIP files or individual documents |
| Formats | ZIP, PDF, DOCX, DOC, ODT, RTF, PPTX, TXT, HTML and Markdown |
| Files in one batch | Up to 300 files |
| Size of each file | Up to 50 MB per document; HTML and Markdown: 4 MB |
| Length of a text | At least 100 words |
| Languages | English, French, German, Spanish, Chinese, Vietnamese and Arabic, with separate evaluation results for each language |
| Folders and names | Each row shows the student folder and filename. Use Check only to filter by either. |
| Scanned PDFs | OCR reads up to 40 scanned pages per file containing printed text in English, French, German, Spanish or Chinese. The result identifies the file as a scan. |
| Not read | Handwriting, image files such as JPG or PNG, scans in other languages and password-protected files |
| Files that cannot be checked | Listed with an explanation; no credits are charged |
| Results | Ordered by the proportion of text flagged and available to download as a CSV file |
Evaluated on student writing
Veread, Fileverity’s own text model, analyzes each text in the batch. Its evaluation uses writing held apart from development, including essays by students and language learners.
Human-written essays
0.30%
of human-written essays by students and language learners were flagged.
AI-written essays
99.1%
of AI-written and AI-assisted essays in the learner-writing evaluation were detected.
Fully AI-written text
98.6%
of fully AI-written texts were detected across seven languages.
Models excluded from development
94.9%
of texts from AI models never used in development were detected.
Held-out validation of Veread 1.1, November 2026Explore evaluation results
How to read a result
Use a result to guide closer reading and a conversation with the student. It cannot establish AI use or misconduct on its own.
How to review a flagged submission
- Open the result, read the highlighted passages in context and compare them with the student’s earlier work.
- Invite the student to explain their argument and share relevant drafts, notes and sources.
- Consider the student’s work across several assignments before drawing a conclusion.
- Follow your institution’s procedures. An unflagged result does not establish that the work was produced without AI assistance.
Who can use it and what it costs
Who can use it
Bulk checks are open to verified teachers, and to organizations and API customers whose workspace has purchased credits. Free, Personal and Pro accounts do not include them on their own.
Free for teachers
A verified teacher’s bulk checks count toward the teacher plan’s allowance of 2,000,000 characters a month. Beyond it, each text uses credits.
Organizations and API customers
Each text uses 2 credits per 1,000 characters or part thereof. Files that cannot be analyzed are not charged.
Schools and developers
Run the same check from your own systems
The text detections API accepts documents as well as plain text, allowing institutions and developers to submit work from a learning platform or an existing workflow.
Frequently asked questions
Can I check a whole class for AI-written text at once?
Yes. Add a ZIP file downloaded from your learning platform or select individual documents. A batch can contain up to 300 files. Each document is analyzed separately, with an assessment, flagged proportion and highlighted passages for completed checks.
How do I download all my students’ submissions?
Use your platform’s own download. In Google Classroom, open the assignment’s folder in Drive and download it. In Moodle, choose Download all submissions. In Canvas, choose Download Submissions. In Blackboard, use Download Submissions in Ultra or Assignment File Download in Original. In Brightspace, select all submissions and choose Download. For Teams, download the student folders from Student Work in the class team’s SharePoint site and use Check only to select the assignment.
Which file types can I upload, and are online-text submissions read?
Upload ZIP archives or individual PDF, DOCX, DOC, ODT, RTF, PPTX, TXT, HTML and Markdown files. Each document can be up to 50 MB, except HTML and Markdown files, which have a 4 MB limit. Online text submissions that your platform exports as HTML files are read. Scanned PDFs are read with OCR when they are printed text in English, French, German, Spanish or Chinese. Handwriting, image files such as JPG or PNG and password-protected files are not read; they are listed with the reason and not charged.
Do I need my school administrator or a plugin?
No plugin or administrative installation is required. Download submissions using your normal platform access and add them to the batch checker in your browser. Continue to follow your institution’s rules for handling student work.
Does it work with Google Classroom, Canvas, Moodle, Teams, Blackboard and Brightspace?
Yes. The batch checker reads supported files from the ZIP archives or folders you download. It runs separately from the learning platform and does not require a direct connection to it.
What does a result mean, and what do I do with a flagged text?
Each completed check returns an assessment: likely written by AI, mixed writing signals or likely written by a person, together with the proportion flagged. Open the result to review the passages in context. Compare them with earlier work and invite the student to explain their reasoning or share drafts. The result alone does not establish AI use or misconduct.
Can it check essays in other languages?
Yes. English, French, German, Spanish, Chinese, Vietnamese and Arabic are evaluated separately. Text analysis requires at least 100 words.
Who can use bulk checks, and what do they cost?
Bulk checks are open to verified teachers, and to organizations and API customers whose workspace has purchased credits. A verified teacher’s bulk checks count toward the teacher plan’s allowance of 2,000,000 characters a month, then use credits. For organizations, each text uses 2 credits per 1,000 characters or part thereof. Free, Personal and Pro accounts do not include bulk checks on their own. Institutions can use education pricing with a shared balance, while developers can run the same analysis through the text detections API.
Guides
Practical guidance for review teams.
What each check establishes, where it stops and which evidence to request next.
Evaluating detection
How to Check a Whole Class Set of Essays for AI at Once
Review class submissions for AI writing in one batch. Download steps for Classroom, Canvas, Moodle, Teams and more, with guidance on interpreting results.
9 min read ·
Evaluating detection
What to Do When an AI Detector Flags a Student's Essay
How teachers can review a flagged essay: read passages in context, examine the writing process, talk with the student and document a fair decision.
11 min read ·
Evaluating detection
Which AI Wrote This? How ChatGPT, Claude, Gemini, Grok, DeepSeek, Qwen and GLM Write
The phrases, punctuation and structure that set seven AI model families apart from human writers, with multipliers, the method and why they are not proof.
6 min read ·
Start with a class set of submissions.
Download submissions from your platform, add the ZIP file to the batch checker and review the results together.