User Agreement

This revision is effective August 30, 2026. Website: neuroslop.net. Service owner: the independent operator of the Neuroslop project; contact: info@neuroslop.net, Telegram **@AITextChecker**. By using the Neuroslop website, you accept these terms.

What the service does

• Text checking: the detector identifies AI patterns and assigns a liveliness score. Checking is free and does not require registration; the checked text is not stored.
• Humanization: a neural network rewrites the text according to our methodology. It is available

after login and consumes the free daily limit or paid credits. Eligible successful website humanizations may enter the protected quality-sample process described below.

• RubricLock: the local structural comparison of an assignment and draft is free. The separate

“Deep Review” runs only after explicit consent, requires a verified account, and consumes three paid credits per request.

• Version comparison: the local comparison is free. The separate Deep Meaning Review runs only

after explicit consent, requires a verified account, and consumes one paid credit per request.

• EvidenceLock: the local comparison of a claim with a supplied source excerpt is free. The separate

deep review runs only after explicit consent, requires a verified account, and consumes one paid credit per request.

Account

• Registration: by email with code confirmation, or via Google or Telegram.
• You are responsible for the security of your password and access to your email.
• One person: one account. Farming free limits through disposable emails and multiple accounts is prohibited; such accounts are blocked without compensation.
• Maximum of 5 active devices per account.

Protected quality samples

While collection is enabled in “Data and privacy,” an eligible successful humanization made by a verified account through our same-origin website may be retained under `humanize-quality-v2` for no more than 30 days. API, bot, unverified-account, cross-origin, sensitive, and detected minor-related texts are excluded. An authorized operator must manually inspect a pair before it can be approved for evaluation or training; a user's quality label never auto-approves it. The account owner can download active samples, turn future collection off and delete all active samples, or turn collection on again. Details are in the Privacy Policy.

The user retains ownership of the text. While protected collection is enabled, the user grants us, for the active retention period, a non-exclusive, royalty-free permission to store, reproduce, analyze, label, and transform an eligible source and result solely to evaluate, test, tune, and train Neuroslop's detection and humanization algorithms. The permission for an active record ends when it is deleted. It does not require us to reverse completed processing or delete aggregate or model improvements from which the source and result cannot be reconstructed.

The user must own the text or have the right to permit this use and must not submit another person's personal data, children's data, confidential information, or third-party material without permission. Automated filters reduce this risk but do not replace the user's responsibility or manual review.

Usage restrictions

• It is prohibited to use the service to create content that violates the law.
• Automated requests to the website bypassing the public API are prohibited.
• Request limits: specified in the interface and may change.

Payment

Paid credits for humanization, RubricLock Deep Review, Deep Meaning Review, and EvidenceLock are purchased in packages; developers can also buy API unit packages (a key's monthly quota). Prices are listed on the tariffs page and the API page. The refund procedure is described in the Refund Policy.

Liability

The service is provided "as is." The detector's score is the algorithm's evaluative judgment; it is not a fact and cannot be the sole basis for decisions about people (termination, student evaluation, etc.). Check the humanization result before publishing: you are responsible for the content of the published text. RubricLock does not predict a grade or prove factual accuracy, authorship, originality, or the absence of plagiarism; you must still compare the result with the assignment yourself. Deep Meaning Review is an editorial heuristic: it may miss a meaning change or flag an acceptable edit. It does not establish factual accuracy; you must review both versions and important conditions yourself. EvidenceLock assesses only how a claim relates to the source excerpt you supply. It does not verify the source's authenticity, completeness, or scientific quality and does not establish that the claim is true; you must review the source and its context yourself.

Changes to terms

We may update the agreement. Continued use after the publication of a new version constitutes acceptance of it.

A material change to the quality-sample purpose, data categories, or retention period applies only prospectively after an updated notice and policy version.

Mandatory rules

This agreement does not limit rights that cannot be waived under applicable law.

Contacts

Email: info@neuroslop.net · support on Telegram: **@AITextChecker · the service is also available as the Telegram bot @AITextChecker_bot** (t.me/AITextChecker_bot).