EU AI Act · GDPR

AI Transparency Notice

Last updated: March 2026

1. Introduction

LegisGate uses artificial intelligence to assist with regulatory compliance analysis, exposure evaluation, and document processing. We are committed to transparency about how AI is used in our Service.

This notice explains which AI systems we use, what data is processed, how decisions are made, and your rights under the EU AI Act and the General Data Protection Regulation (GDPR). It supplements our Privacy Policy and Terms of Service.

2. AI Systems Used

  • Provider: Anthropic (Claude). All AI processing is inference-only — your data is never used to train AI models.
  • Purpose: Non-regulatory prose only — executive narrative summaries, vendor documentation research summaries, optional conversational features, and similar machine-assisted research aids. Regulatory findings are never AI-authored; they come from the intelligence library SSOT and deterministic rules engine.
  • EU AI Act classification: Limited Risk (Article 50). Transparency obligations apply, and this notice fulfills those requirements.

3. How AI is Used in LegisGate™

LegisGate™'s compliance intelligence engine (patent pending) is grounded in the LegisGate Meridian™ and deterministic regulatory rules. Regulatory findings are never AI-authored. Anthropic Claude is used only for live vendor documentation research summaries and optional conversational features — not executive narrative, findings, or obligations.

LegisGate Compass™ Reports

Regulatory findings — obligation text, citations, severity tied to regulatory substance, and EU AI Act classification — are produced by Validated Templates from the intelligence library SSOT or Deterministic Rules applied to your confirmed intake. AI assists only with executive narrative sections and vendor documentation research summaries. Risk scores (0–100) are directional indicators. Your Data Protection Team reviews outputs and records deployment decisions — LegisGate™ does not make compliance determinations on your behalf.

Ask Meridian™ Assistant

A conversational interface that uses AI to answer regulatory questions in your organizational context. Users are informed they are interacting with an AI-backed feature. Responses are informational only and do not constitute legal advice.

Regulatory Intelligence Summaries

Machine-assisted summarization of regulatory updates from official sources, with impact framing relevant to your organization. Summaries are labeled as research aids, not primary legal authority or SSOT finding substance.

Privacy Policy Analyzer

Machine-assisted extraction of structured information from vendor privacy policies, including data collection practices, retention periods, and third-party sharing. Extracted facts feed into report narrative and review workflows — not into SSOT regulatory finding templates.

Document Processing

Machine-assisted classification and extraction from uploaded documents such as Data Processing Agreements (DPAs), SOC 2 reports, and vendor security questionnaires. Outputs support human review; they do not replace SSOT regulatory findings.

EU AI Act Classification

EU AI Act classification follows a structured Art. 6 / Annex III sequence executed by the deterministic rules engine. Classifications are indicative and subject to human review.

Breach Notification Drafting

Machine-assisted drafting of breach notification letters based on incident details and applicable regulatory requirements. All drafts require human review and editing before use — they are starting points, not final documents.

4. Data Processing for AI

What data is sent to the AI provider

Report intake details, vendor documentation text, executive narrative inputs, and user questions submitted through optional AI-backed features. Regulatory finding substance from the intelligence library SSOT is not generated by the AI provider.

What is never sent

Passwords, authentication tokens, payment information, and raw personal data of data subjects are never transmitted to the AI provider.

Legal basis (GDPR)

  • Legitimate interest (Article 6(1)(f)) — for B2B regulatory compliance analysis where AI processing is necessary to provide meaningful compliance insights.
  • Contract performance (Article 6(1)(b)) — where optional AI-backed narrative features are part of the contracted Service.

Data retention

AI inputs and outputs are stored in your organization's LegisGate account and subject to our standard data retention policies. Anthropic does not retain data beyond the API request per their data processing terms.

Sub-processor

Anthropic PBC, San Francisco, CA, USA. Standard Contractual Clauses (SCCs) are in place for transfers of personal data from the EEA/UK to the United States.

5. Human Oversight

  • Regulatory findings in LegisGate Compass™ Reports are SSOT-backed and never AI-authored.
  • LegisGate Compass™ Reports require human review before deployment decisions are recorded.
  • Risk scores are directional indicators, not definitive determinations.
  • Users can override, edit, or reject machine-assisted narrative content at any stage.
  • The workflow includes mandatory human review stages before a report can be finalized.
  • Where outputs are substantially machine-assisted, they are labeled appropriately in the Service.

6. Your Rights (GDPR Article 22)

Under the GDPR and the EU AI Act, you have the following rights in relation to machine-assisted processing in the Service:

  • Human intervention: The right to obtain human intervention in decisions significantly affected by automated processing.
  • Express your view: The right to express your point of view and to contest machine-assisted outputs used in your workflow.
  • Explanation: The right to an explanation of the logic involved in automated processing that affects you.
  • Opt out: The right to opt out of optional AI-backed narrative features by contacting our support team.
  • Access, rectification, and erasure: The right to access, rectify, and erase data that has been processed by AI systems.
  • Supervisory authority: The right to lodge a complaint with your data protection supervisory authority.

To exercise any of these rights, contact us at privacy@legisgate.com.

7. AI Limitations

  • Machine-assisted narrative outputs may contain errors or outdated information.
  • AI narrative is based on information available at the time of processing and may not reflect the most recent regulatory changes.
  • Regulatory landscapes change frequently; all outputs should be verified against current law.
  • AI cannot replace qualified legal, compliance, or privacy professionals. LegisGate is a tool to assist — not substitute — professional judgment.

LegisGate™ LegisGate Compass™ Reports combine the LegisGate Meridian™ — a curated regulatory intelligence reference — deterministic regulatory rules, and citations verified to primary statutory source; regulatory findings are never AI-authored, and AI assists with live vendor documentation research only.

8. Contact

For questions about AI processing or to exercise your rights:

9. Updates

This notice may be updated as our AI capabilities evolve or as regulatory requirements change. We will notify you of material changes via the Service or email. The "Last updated" date at the top indicates when this notice was last revised.

Talk to usWe're here to help
AI Transparency Notice - Where We Use AI