Terms of Service
Last updated: March 7, 2026
1. Acceptance of Terms
By accessing and using the Pauhu Data Marketplace ("Service"), you accept and agree to be bound by these Terms of Service. If you do not agree, do not use the Service.
2. Description of Service
Pauhu Data Marketplace (pauhu.eu) provides access to annotated EU institutional data feeds in all 24 official EU languages. Data sources include:
- CURIA — Court of Justice case law
- ECB — European Central Bank statistical data
- ECHA — Chemical substances registry (REACH, CLP)
- EMA — Medicinal products data and EPARs
- EPO — European patent publications
- Parliament — Legislative proceedings and votes
- EUR-Lex — EU law, regulations, and directives
- Eurostat — EU statistical indicators
- IATE — EU terminology database (2.4M terms)
- Legislative Observatory — Legislative procedure tracking
- TED — Public procurement tenders
- Wikidata — EU entity knowledge base
Data is annotated across 21 EuroVoc industry domains and delivered via REST API or Eclipse Dataspace Connector (LDS).
3. Subscription Tiers
The Service is available in four consumption-based volume tiers:
- S1 Pay-as-you-go: €33.49 per 1,000 API calls. All 20 EU sources, 27 national law databases, all 24 languages. REST API (JSON, CSV, XML). Email support.
- Volume C2: €1,689/month. 50,000 API calls included, overage at €27.53/1K calls. Webhooks, annotations, semantic search. SLA: 99.5%.
- Volume C3: €4,933/month. 250,000 API calls included, overage at €20.09/1K calls. Full EuroVoc + CPV annotations, IATE terminology export, 290K cross-references. SLA: 99.9%.
- Data Container (C4): €36,995/month. 2.5M API calls included, overage at €15.07/1K calls. Self-hosted via Docker and Eclipse Dataspace Connector (LDS). Air-gap capable. Unlimited users per instance. SLA: 99.99%.
All tiers include access to all 20 EU sources, 27 national law databases, and all 21 EuroVoc domains. All prices are MACC-eligible via Azure Marketplace.
4. License Grant
Upon purchase, you are granted a non-exclusive, non-transferable license to use the data feeds for:
- Internal use, research, and application development
- Integration into your products and services
Alignment Corpora subscribers are additionally licensed for:
- AI/ML model training using the complete multilingual parallel structure
- Fine-tuning on aligned parallel text across all 24 EU languages
- RAG (retrieval-augmented generation) and inference serving
- Commercial derivative works that preserve multilingual alignment
Language justice clause: Alignment Corpora may not be used to extract, isolate, or preferentially train on a single language or subset of languages to the exclusion of others. The parallel structure across all 24 official EU languages is integral to the licensed product.
This requirement reflects:
- The Charter of Fundamental Rights of the European Union: Article 22 (respect for linguistic diversity) and Article 21 (prohibition of discrimination based on language), which have the same legal value as the EU Treaties since the Treaty of Lisbon
- The equal legal authority of all 24 official EU language versions established by the EU Treaties and Regulation No 1/1958
- The linguistic rights framework of the Finnish Constitution (Section 17) and the Language Act (423/2003)
- The follow-up indicators for linguistic rights developed by the Finnish Ministry of Justice (Publication 35/2018, OMSO 35/2018, ISBN 978-952-259-714-4), which establish structural, process, and outcome indicators for monitoring the realisation of linguistic rights
Pauhu’s three annotation layers — EuroVoc domain classification (structural), deontic modality tagging (process), and CPV procurement codes (outcome) — are modelled on this indicator framework. The data architecture is designed to ensure that no language is treated as subordinate in AI training applications.
You may not redistribute the raw data feeds to third parties as a competing data service without explicit written permission.
5. Data Sources and Annotation Framework
All datasets are derived from publicly available EU institutional sources. Pauhu adds value through:
- Data cleaning and normalization across 24 languages
- Multilingual alignment and cross-referencing (290,000 directive-to-national-law links from EUR-Lex Sector 7)
- Structural layer: EuroVoc domain classification (21 domains from the EU’s official thesaurus) — identifying what area of law exists
- Process layer: Deontic modality tagging (obligation, prohibition, permission, exemption) — identifying what the law requires
- Outcome layer: CPV procurement code classification (EU standard) — identifying where law meets real-world activity
- IATE terminology alignment (2.4M terms across up to 24 languages per concept)
- API access and sovereign delivery infrastructure (REST API and Eclipse Dataspace Connector)
This three-layer annotation framework corresponds to the structural, process, and outcome indicators defined in the UN human rights indicator framework, as applied to linguistic rights by the Finnish Ministry of Justice (OMSO 35/2018).
6. Payment Terms
Subscriptions: Billed annually or monthly per EuroVoc domain. Cancel anytime with 30 days notice.
Free trial: 30-day evaluation with rate-limited access to all sources.
All prices in EUR. VAT applied where required by EU law.
7. Data Accuracy
While we strive for accuracy, data feeds are provided "as is". We do not guarantee 100% accuracy of annotations or classifications. Users should validate data for their specific use cases.
8. Prohibited Uses
You shall not use the Service, any data feeds, annotations, or outputs for:
- Surveillance, tracking, or monitoring of individuals, including locating, profiling, or identifying natural persons for intelligence, law enforcement, or military purposes
- Development or operation of weapons systems, military targeting, or lethal autonomous systems
- Mass surveillance, social scoring, or biometric identification in public spaces
- Any purpose prohibited by the EU AI Act (Regulation (EU) 2024/1689) Article 5, including subliminal manipulation, exploitation of vulnerabilities, and real-time remote biometric identification in publicly accessible spaces
- Any purpose that violates fundamental rights as recognised by the EU Charter of Fundamental Rights
Violation of this section constitutes a material breach entitling Pauhu to immediate termination without cure period.
9. AI Transparency (EU AI Act Art. 52)
In accordance with the EU AI Act (Regulation (EU) 2024/1689) Article 52, we disclose the following:
- The Service uses AI systems for data annotation, deontic classification, topic classification, and semantic search
- Annotations are probabilistic classifications, not legal advice — confidence scores are provided where applicable
- All AI inference runs browser-native (ONNX) with no server-side processing of user queries beyond retrieval
- No emotion recognition, biometric categorisation, or social scoring systems are used
For questions about AI systems used in the Service, contact: legal@pauhu.eu
10. Digital Services Act (DSA) — Point of Contact
In accordance with the Digital Services Act (Regulation (EU) 2022/2065), the following information is provided:
- Point of contact (Art. 11): legal@pauhu.eu
- Legal representative: Pauhu Ltd (Y-tunnus: 0768171-8), P.O. Box 292, 00101 Helsinki, Finland
- Languages: English, Finnish, Swedish
Users may report illegal content or submit complaints via legal@pauhu.eu. We will acknowledge reports within 24 hours and respond substantively within 7 business days.
11. Limitation of Liability
Pauhu's liability is limited to the amount paid for the Service in the 12 months preceding any claim. We are not liable for indirect, incidental, or consequential damages.
12. Governing Law
These terms are governed by Finnish law. Disputes shall be resolved in the courts of Helsinki, Finland.
13. .eu Domain
The .eu top-level domain is established by Regulation (EC) No 733/2002 of the European Parliament and of the Council. Pauhu Ltd is eligible to operate under .eu as an undertaking with its registered office in Finland, a Member State of the European Union (Article 4(2)(b)).
14. Linguistic Rights
The Charter of Fundamental Rights of the European Union establishes linguistic diversity as a fundamental right. Article 22 requires the Union to respect cultural, religious, and linguistic diversity. Article 21 prohibits discrimination on grounds of language. These provisions have the same legal value as the EU Treaties since the entry into force of the Treaty of Lisbon (2009).
EU citizens have the right to communicate with EU institutions in any of the 24 official languages and to receive a reply in the same language. All EU legislation is published in all 24 official languages, and each language version is equally legally authoritative (Regulation No 1/1958, as amended).
Pauhu Ltd is incorporated in Finland, a jurisdiction with constitutional protection for linguistic rights (Section 17 of the Constitution of Finland). The Language Act (423/2003) establishes the right to use Finnish and Swedish before authorities. The Sámi Language Act (1086/2003) protects Sámi linguistic rights in the Sámi Homeland. The European Charter for Regional or Minority Languages further protects linguistic diversity at the Council of Europe level.
The Finnish Ministry of Justice monitors the realisation of linguistic rights through follow-up indicators based on the United Nations human rights indicator framework, as published in Follow-up Indicators for Linguistic Rights (Ministry of Justice, Finland, Publication 35/2018, ISBN 978-952-259-714-4). These indicators measure linguistic rights across three dimensions: structural (legal instruments), process (policy implementation), and outcome (experiences of rights-holders).
Pauhu’s data architecture and language justice clause are informed by this framework. We treat all 24 official EU languages as equal in data structure, metadata application, and access. No language is primary; no language is derivative.
15. Contact
Questions about these terms: legal@pauhu.eu
Pauhu Ltd
Helsinki, Finland
EU jurisdiction