Article-level citations for compliance, legal, and security work — in the AI client your team already uses. You bring the AI. We bring the regulations, frameworks, and case law.
Security engineering, GRC and privacy, and legal teams each get the workflow shape they expect — on a shared corpus of regulation, standards, and your own internal documents.
DORA, NIS2, EU AI Act, and DPIA workflows that surface gaps with article-level citations and exportable audit trails. Structured workflow tools are Team tier (waitlist); Free and Premium answer the same questions as search-grounded citation work in your own AI client.
Legislation, case law, and preparatory works — current text, source URLs, freshness dates, with private documents and customer evidence in the same citation contract.
STRIDE, LINDDUN, ISO/SAE 21434 TARA, and OWASP ASVS reviews — findings mapped to ASVS, ISO 27001, and the regulations that make them urgent.
We do not answer without sources. We do not hide failed lookups. The product is the audit trail.
Threat models, gap analyses, and legal questions that span standards, regulations, and internal policies — answered across jurisdictions in one pass, with every claim tied to the exact source.
Every answer carries its sources. Every workflow carries its trace. When a source is unavailable, we show it — we do not invent.
One MCP connection (Model Context Protocol — the open standard that lets AI assistants call external tools). Use Claude, Cursor, VS Code, GitHub Copilot, Open WebUI — any client good at tool-calling. Same evidence trail whether the task starts in security engineering, GRC, privacy, or legal.
Six of the questions security, compliance, and legal teams bring to Ansvar on any given Tuesday. Every answer ships with its sources. On Free and Premium, your own AI client runs these as search-grounded citation work; structured workflow tools with exportable audit trails are Team and Company tier (waitlist).
Fixed-scope consulting engagements from €2,000 — threat models, DPIAs, and gap analyses delivered by Ansvar, scoped on a call. See services.
Each assertion in a threat model, gap analysis, or research answer is traced back to the exact page and paragraph — with confidence scores and direct links.
Every answer is source-backed — with jurisdiction, article, and source URL.
Every workflow is traceable — tool calls, data sources, and confidence scores are persisted.
Workflow results are exportable — audit-ready metadata for enterprise review (workflow tools: Team and Company tier).
Failed sources are visible — if a data source is unavailable, we say so; we do not fall back to AI guesswork.
Citations are validated — a deterministic pipeline cross-checks every reference before the answer renders.
Ansvar connects through MCP, without becoming another chatbot. The gateway handles retrieval and citation validation behind the scenes; the conversation stays in your AI client.
EU-hosted, no model proxy, open connector code. The things your procurement team asks about on call #2.
Hosted in the EU on Hetzner. Cloudflare provides edge and TLS termination under SCCs. GDPR-compliant processor agreements available.
Ansvar runs no server-side AI model. Inference flows directly from your AI client to your model provider — we never see or proxy it.
Open Law connectors ship under Apache 2.0 — inspect how each corpus is parsed and sourced, and reuse them in your own stack. The hosted gateway runtime is not open source.
Building an agent? Connect any MCP client to gateway.ansvar.eu/mcp — same OAuth, same citation contract. See setup.
Run your first threat model, gap analysis, or legal research session in the AI client you already use. Every answer comes back with its sources.