DetectorScope
DetectorScope is a Python package and demo for building and using auditable AI-text detectors. It explains predictions with named text features, their contributions, and their locations in the input.
What it does
Developers and researchers can detect AI-generated text, inspect why a detector made its prediction, or convert a black-box detector into an auditable one.
- Availability
- PyPI package, Hugging Face demo and downloadable detector models
- Pricing model
- Not established
- Deployment
- Demo on Hugging Face Spaces; package can run locally
- Privacy
- Not established
- Licence
- Apache License 2.0 for software; CC-BY 4.0 for other materials
Capability
Modality
Access
Deployment
Pricing
Target user
Evidence
6 claims- DetectorScope reports named features that contribute to an AI-text prediction. [1]
- The detector can show where features are located in the text. [1]
It can tell you which properties of a text made it predict "AI" and where in the text they are located.
GitHub new repositories. Quoted text is outside CC BY 4.0 and keeps its original rights. - The package can convert a black-box detector into an auditable detector. [1]
* 🔁 **Convert your own black-box detector** into an auditable one by using our pipeline
GitHub new repositories. Quoted text is outside CC BY 4.0 and keeps its original rights. - The package is installable from PyPI as detectorscope. [1]
- Detector models are available for download from Hugging Face. [1]
- The authors describe the detectors as research artifacts and warn against using them for decisions about individuals. [1]
Sources
1
1
GitHub new repositoriesOwner-controlled
https://github.com/google-deepmind/detectorscope
Retrieved 2026-10-29 00:30 UTC
Verified
Versions
1| Version | Effective | Change | Changed fields | State |
|---|---|---|---|---|
| 1 | 2026-10-29 00:30 UTC | publication | None | Current |
Original catalog content is licensed under CC BY 4.0, attributed to wowsuchai. Source quotations and linked content keep their original rights.