OpenPrivacyPolicyTaxonomy/oppt-privacy-policies
Viewer
•
Updated
•
3.65k
•
7
Open Privacy Policy Taxonomy (OPPT)
non-profit
AI & ML interests
Open Privacy Policy Taxonomy (OPPT) is a research initiative creating open, annotated datasets for privacy policy analysis, dark pattern detection, and NLP research. We extend the foundational OPP-115 corpus (Wilson et al., ACL 2016) with modern regulatory categories reflecting GDPR, CCPA/CPRA, and emerging AI governance frameworks. Current Release: OPPT-T1_C1.0_Section_Jan2026 • 123 privacy policies from major tech, finance, healthcare, and data broker companies • 3,651 annotated segments with 14 taxonomy categories • Three-model consensus methodology (Claude Haiku 4.5, GPT-5.2, Gemini-3-flash-preview) • CC BY-NC 4.0 license (commercial licensing available) Paper: https://arxiv.org/abs/2601.20792 Research Focus: • Privacy policy classification and NLP • Dark pattern detection in privacy disclosures • Regulatory compliance analysis (GDPR, CCPA, state privacy laws) • Longitudinal policy tracking Contact: tebrackin@outlook.com (commercial licensing inquiries)
Recent Activity
View all activity
Organization Card
Open Privacy Policy Taxonomy (OPPT)
Advancing privacy policy transparency through open taxonomies and annotated datasets
Mission
We develop standardized frameworks and tools for analyzing privacy policies, enabling researchers and practitioners to:
- Identify dark patterns and deceptive practices in privacy disclosures
- Compare privacy practices across companies and sectors
- Build machine learning models for automated policy analysis
- Track regulatory compliance across jurisdictions
Projects
OPPT v1.0 Taxonomy
A comprehensive 14-category taxonomy for privacy policy classification, extending the foundational OPP-115 scheme with modern regulatory categories (GDPR, CCPA/CPRA, AI Act).
OPPT-T1_C1.0_Section_Jan2026 Dataset
123 privacy policies from major technology companies annotated using OPPT v1.0, featuring:
- 3,651 annotated segments
- Three-model consensus methodology (Claude Haiku 4.5, GPT-5.2, Gemini-3-flash-preview)
- Rich attribute schemas
- Modern regulatory landscape (January 2026)
Paper: Jurisdiction as Concealment (arXiv:2601.20792)
Research Areas
- Privacy policy analysis and NLP
- Dark pattern detection
- Regulatory compliance automation
- Consumer privacy protection
Contact
- Inquiries: tebrackin@outlook.com
- Commercial licensing: tebrackin@outlook.com
Building tools for a more transparent digital privacy landscape
models
0
None public yet