# Ansvar AI > Cited legal evidence and reviewable compliance work, in your AI tools. EU-hosted. Bring your own model. Ansvar AI connects AI clients (Claude, ChatGPT, Copilot, Cursor, your own agents) to verified law, SIS-licensed standards, security intelligence and customer evidence through one MCP gateway. Four products share that connection: legal and regulatory evidence, living security architecture, compliance assessments, and security risk and threat intelligence. Every served row carries its publisher, licence and source URL; a question the served corpora cannot answer is refused, never filled from model memory. Ansvar does not train models on customer data. Ansvar AI is built and operated by Ansvar Systems AB in Sweden. The service is EU-hosted and runs no server-side AI model: model traffic flows from your AI client to your model provider, and your provider bills the tokens while Ansvar bills the connection. ## Product - [Legal and regulatory evidence](https://ansvar.eu/evidence): law, regulation, standards, case law, preparatory works and agency guidance across Europe, the UK, the US and Canada, every row cited to the publisher that issued it. Free answers one jurisdiction or framework per question; Solo reaches the whole served fleet; Premium adds case law, preparatory works and guidance. - [Living security architecture](https://ansvar.eu/architecture): a model of components, data flows and trust zones that your agents keep current through proposals your people approve, with every fact cited to where it was seen. Runs as a workspace you host yourself; 30-day evaluation, then a licence. - [Compliance assessments](https://ansvar.eu/assessments): gap analyses, DPIAs and FRIAs, conformity and authorisation runs your own agent walks end to end on a system you describe, or on your own documents from Team. Every finding cites its provision or is marked unresolved; every report carries a delivery receipt. - [Security risk and threat intelligence](https://ansvar.eu/cyber-security): CVE, KEV, EPSS, exploit and CISA ICS advisory lookups on every plan; MITRE ATT&CK and ATLAS on every plan, CAPEC, CWE and D3FEND from Premium; threat models and TARAs your agent runs; from Premium a vulnerability assessment, an ICS advisory run or a deferral dossier scores each finding against the system you described; from Team your own scoring contexts, recorded review decisions and OpenVEX export. - [ISMS as Code](https://ansvar.eu/isms-as-code): add-on on an annual contract. The management system as a repository your agent operates and your people approve; ISO 27001 or NIST-keyed; paid pilot credited on conversion. - [Deploy](https://ansvar.eu/deploy): hosted by Ansvar in the EU (available), or the customer plane in your Kubernetes and on-prem or air-gapped delivery (design-partner pilots). - [Examples](https://ansvar.eu/examples): complete sample deliverables and captured production runs, each published verbatim on its own page — a NIS2 gap analysis with the policy attached (licensed ISO 27001 clauses from SIS fetched in the loop), a security-policy gap analysis, DPIA for an HR vendor, STRIDE threat model, EU AI Act classification, and a paragraph-cited document review — with the cited answers exactly as the gateway returned them. - [Solutions](https://ansvar.eu/solutions): three entry points over the same gateway: legal research and casework, enterprise security and compliance assurance, plus product cybersecurity and regulated engineering. Sector pages remain under /sectors/. - [Financial services](https://ansvar.eu/solutions/financial-services): served EU and national financial regulation, ECB and EIOPA material, and ESMA sanctions register rows. Statute research on Free, interpretive research on Premium, and DORA assessments with uploaded evidence on Team. Coverage and plan access vary by source. - [Legal research](https://ansvar.eu/solutions/legal-research): search served legislation and, on Premium and above, available case law, preparatory works, and agency guidance in your chosen AI client. Results retain publisher links and unresolved evidence gaps. - [Pricing](https://ansvar.eu/pricing): Free €0 (one jurisdiction or framework per question, 100 searches a day, public CVE/KEV/EPSS intelligence, one assessment a month on a described system, watermarked; B2B, no card) · Solo €29/month or €290/year (every jurisdiction in one question, 750 searches a day, two assessments a month, watermarked; the seat may be an agent) · Premium €249/month or €2,490/year per seat (case law, preparatory works and guidance; the assessment catalogue on a described system, five runs a month, JSON plus unwatermarked HTML, PDF and DOCX reports; CAPEC/CWE/D3FEND enrichment; run-scoped effective-risk analysis inside a vulnerability assessment, an ICS advisory run or a deferral dossier; 5,000 searches a day) · Team €490/month or €4,900/year per seat (each seat is a person or an agent credential; five service credentials per workspace included without taking a seat; customer-document grounding cited by paragraph, one architecture workspace licence, standalone effective-risk tools with persistent scoring contexts, review decisions and OpenVEX export, twenty runs a month per seat pooled across the workspace then €35 a run, 50,000 searches a day pooled) · Company from €2,000/month billed annually on contract (signed audit ledger per tenant, five architecture licences and the plane, 100 assessments a month, 500,000 searches a day pooled). Annual is ten months on self-serve plans. A seat is a seat whether a person or an agent holds it. Premium, Team and Company admins mint org-owned OAuth2 client_credentials for headless agents; each receives the same gateway tool surface as a signed-in seat on the plan. On Premium each credential takes a full seat; on Team and Company five per workspace take no seat and each further credential takes one. Full quotas and the refusal contract: /limits. - [Security](https://ansvar.eu/security): EU-hosted and dedicated deployment options, the customer-managed design-partner path, model/data flows, operational logging, and Company signed receipts. - [Coverage](https://ansvar.eu/coverage): the live audited jurisdiction and corpus inventory — legal jurisdictions, security frameworks, and regulatory domains, with provision counts and per-country pages. - [Standards](https://ansvar.eu/standards): clause and control text of ISO/IEC 27001, 27002, 27005, 42001 and ISO/SAE 21434 served through the gateway under a SIS license, cited rather than paraphrased. Optional per-seat add-on; buyers pick their standards at checkout. A named person can hold a standards seat on any plan, including Free. A machine holder requires a Team or Company service credential. Terms: /standards-addon-terms. - [Workflows](https://ansvar.eu/workflows): the directory of served assessment families. Premium covers the catalogue on a described system, with unwatermarked HTML, PDF and DOCX exports (a few variants start at Team); Team adds families that require customer documents. `list_workflow_types` returns the live types available to the signed-in account. - [Regulatory gap analysis](https://ansvar.eu/workflows/gap-analysis): where you stand against a regulation, requirement by requirement, each finding cited to its provision or marked unresolved. One assessment spine with a variant per regime — NIS2 (incl. the Dutch transposition), DORA, CRA, EU AI Act, MDR/IVDR/MDCG, UNECE R155, the Machinery Regulation and the two drone regimes. Startable on the free monthly run for the generic, NIS2, DORA, CRA and AI Act variants. - [TARA workflows](https://ansvar.eu/cyber-security#catalogue): staged threat analysis and risk assessment for vehicles, OT, rail, robots and drones — risks banded on NIST 800-30 scales, every regulatory anchor fetched and cited or marked unresolved. Sector variants over one enterprise risk spine (ISO/SAE 21434, IEC 62443, CLC/TS 50701, UN R155). Premium and above. - [Canonical control library](https://ansvar.eu/control-library): one NIST 800-53r5 control spine under CSF 2.0, with pointer mappings to other frameworks typed by set-theory relationship (equivalent, superset-of, subset-of, intersects-with, related-to). Outside the generated 800-53 identity map, coverage is claimed only where a human has reviewed the typing — the NIST-published crosswalks we ingest arrive untyped and claim nothing. Seven tools, Team and Company. - [Services](https://ansvar.eu/services): Fixed-scope consulting engagements on the Ansvar gateway for AI Act readiness, gap analysis, and custom workflow work. Paid deliveries are citation-grounded, senior-reviewed, and fixed-scope. The four fictional sample documents below show the report structure and state that senior review is pending; the deferral dossier is a captured production-workflow sample. - [Product security and engineering](https://ansvar.eu/solutions/product-security): for teams who ship software or connected products, including apps built with an AI coding agent. Connect the same agent to run a STRIDE threat model, a LINDDUN privacy analysis, and live CVE/KEV/EPSS dependency checks. Factual findings cite the served law or advisory, or remain unresolved. - [Sample threat model](https://ansvar.eu/services/sample-threat-model): a worked STRIDE threat-model sample for a fictional LLM-backed returns API, labelled senior review pending. - [Sample DPIA](https://ansvar.eu/services/sample-dpia): a worked GDPR Art. 35 assessment for a fictional workplace-wellness app, labelled senior review pending. - [Sample gap analysis](https://ansvar.eu/services/sample-gap-analysis): a worked NIS2 gap analysis for a fictional logistics SaaS, labelled senior review pending. - [Sample AI Act readiness](https://ansvar.eu/services/sample-ai-act-readiness): a worked EU AI Act readiness assessment for a fictional HR-tech provider, labelled senior review pending. - [Sample deferral dossier](https://ansvar.eu/services/sample-deferral-dossier): a CRA-era vulnerability deferral dossier captured from a production workflow run, with KEV and Art. 14 flags on real CVEs and fetched CRA/NIS2 pin-cites. - [NIS2 gap analysis template](https://ansvar.eu/templates/nis2-gap-analysis): Free ungated XLSX/PDF — one row per Article 21(2) measure, Article 23 incident-reporting tracker, verdict dropdowns. - [DPIA template](https://ansvar.eu/templates/dpia): Free ungated XLSX/PDF — Article 35 screening with the WP248 criteria, risk register, Article 36 determination. - [EU AI Act high-risk checker](https://ansvar.eu/tools/ai-act-checker): Free interactive classification walk — Art. 5 screen, Art. 6 gates, Annex III, profiling override — cited to the Act. - [NIS2 scope checker](https://ansvar.eu/tools/nis2-checker): Free interactive scope walk — Annex I/II sectors, size thresholds, size-independent categories, essential vs important — cited to Directive (EU) 2022/2555. - [NIS2 explained](https://ansvar.eu/regulations/nis2): NIS2 scope, the essential/important split, the ten Article 21 measures, the Article 23 reporting clock, fines, and national transposition — each claim cited to the directive. - [CRA explained](https://ansvar.eu/regulations/cra): Cyber Resilience Act scope (the Article 2 connectivity test and its carve-outs), the Annex I Part I product properties and Part II vulnerability handling incl. the software bill of materials, the five-year support period, both 24/72-hour reporting clocks, the Annex III class I/II and Annex IV conformity routes, and the three fine ceilings — each claim cited to Regulation (EU) 2024/2847, applicable from 11 December 2027 with Article 14 reporting from 11 September 2026. - [DORA explained](https://ansvar.eu/regulations/dora): DORA scope (the twenty Article 2 entity types + ICT providers), the five pillars, the three-report incident structure, TLPT, the Article 30 contract clauses, the register of information, and the two enforcement tracks — each claim cited to Regulation (EU) 2022/2554, directly applicable since 17 January 2025. - [GDPR explained](https://ansvar.eu/regulations/gdpr): GDPR scope (the Article 2/3 material and territorial tests), the Article 5 principles, the six Article 6 lawful bases and Article 7 consent conditions, the Article 9 special categories, the Chapter III rights on the Article 12 one-month clock, the Article 25/28/30/32 accountability duties, the Article 33/34 breach clocks incl. the 72-hour notification, DPIAs and DPOs, the Chapter V transfer ladder, and the Article 83 fine tiers — each claim cited to Regulation (EU) 2016/679, applicable since 25 May 2018. - [Machinery Regulation explained](https://ansvar.eu/regulations/machinery): EU Machinery Regulation scope (machinery, five related product groups, partly completed machinery, the Article 2 exclusions), the Article 10 manufacturer duties incl. source-code access and digital instructions, the Annex III essential health and safety requirements incl. the 1.1.9 protection-against-corruption and 1.2.1 control-system cybersecurity points, the Annex I Part A/B conformity routes incl. machine-learning safety components, the Digital Omnibus AI Act coupling, and the staged dates — each claim cited to Regulation (EU) 2023/1230 (and (EU) 2026/1744 for the amendments), applicable from 20 January 2027. - [DSGVO- und KI-VO-Compliance: Anbieter im Vergleich](https://ansvar.eu/de/dsgvo-ki-vo-compliance-anbieter): German-language provider comparison — which offerings deliver GDPR and EU AI Act compliance with audit-ready, source-cited reports. Four solution categories (GRC platforms, data-protection tools, AI-governance vendors, source-cited evidence layers that run inside the customer's own AI client) compared on six axes: ongoing monitoring, cited audit-ready reports, legal-source currency, AI Act coverage, EU hosting, integration path. Categories, not named vendors. - [DSGVO- und KI-VO-Compliance mit auditfähigen Nachweisen](https://ansvar.eu/de/dsgvo-compliance): German-language landing page — GDPR and EU AI Act compliance inside your own AI client: article-level citations, audit-ready DPIA and gap-analysis reports, current legal sources, EU-hosted. - [Ansvarsfull AI](https://ansvar.eu/sv/ansvarsfull-ai): Swedish-language landing page — cited answers for Swedish compliance teams on the EU AI Act, GDPR, and ISO 42001, with article-level citations from public sources, EU-hosted. - [面向欧洲市场的产品安全与合规评估](https://ansvar.eu/zh-cn): Simplified Chinese introduction for Chinese product teams entering Europe. EU regulations and served Chinese national laws with original-source citations, threat modeling, TARA and gap analysis. Chinese local, judicial and departmental layers and GB/GB-T body text are excluded. - [AVG-naleving](https://ansvar.eu/nl/avg-naleving): Dutch-language landing page — article-level citations for Dutch compliance teams on the AVG (GDPR), EU AI Act, and ISO 42001, from public regulatory sources, EU-hosted. - Data access: programmatic access for developers is MCP-only via gateway.ansvar.eu over OAuth 2.1 — there is no REST data API. Client setup is under Docs: Quickstart below. - [Rate limits](https://ansvar.eu/limits): Per-tier limits reference for the gateway — rate limits, concurrency caps, the Free-tier daily search quota, and the error contract (caps surface as JSON-RPC -32000 `cap_exceeded` inside the MCP session, not HTTP 429; a known tool outside your plan returns a refusal naming the plan that includes it; unknown tool names return -32601). Includes sample retry code. - [How Ansvar AI operates](https://ansvar.eu/trust): Operating posture — your AI client talks to your model provider directly, deterministic citation validation, EU-hosted retrieval, and no training on customer data. ## Agent-readable surfaces - [Full site content](https://ansvar.eu/llms-full.txt): every indexable page's title, description, and main-content text in one file — build-generated from the same prerender the pages serve, so it cannot lag them. - [Evaluate](https://ansvar.eu/evaluate): hand your agent a self-contained evaluation pack — it interviews the user, checks their jurisdictions and frameworks against the embedded audited coverage snapshot, and delivers a fit verdict bound to four honest outcomes including "no current fit". The pack itself is served at [/agent-evaluation.md](https://ansvar.eu/agent-evaluation.md) (format-versioned copy at /agent-evaluation-v1.md; both regenerate as coverage moves, and each pack carries a content fingerprint identifying its exact text). - [Claude connector directory listing](https://claude.ai/directory/connectors/ae4b5f7b-6a64-4282-a611-86498a8bf7a0): one-click connect for claude.ai and Claude Desktop users (Claude login required to view); the listing also carries the Claude Code and Claude API snippets. - [Docs: What Ansvar does](https://ansvar.eu/docs/overview): the capability map — each capability area (law research, case law and guidance, claim checking, workflows, documents, org standards, regulatory intelligence, vulnerability intelligence, effective risk, control library, EU regulation engines, audit ledger), what it returns, which plan includes it, and its guide. - [Docs: Quickstart](https://ansvar.eu/docs/quickstart): connect an AI client to the gateway and run a first cited query. - [Docs: Agent skills](https://ansvar.eu/docs/agent-skills): downloadable SKILL.md files — the Using Ansvar routing skill (install once; standing instructions that route bare compliance questions through the connector instead of answering from model memory; ready-made ZIP at https://ansvar.eu/skills/using-ansvar.zip) plus the task skills that build on it (CRA vulnerability obligations; EU incident reporting navigator; regulatory threat model; ISO standards expert) and the agent-affect engineering-practice pack (whine/cringe/protect end-of-task findings, no account needed) — CC BY 4.0. - [Docs: MCP basics](https://ansvar.eu/docs/concepts/mcp-basics): what MCP is and how the gateway uses it — for readers new to the protocol. - [Docs: Connect ChatGPT](https://ansvar.eu/docs/setup/chatgpt): a custom connector in ChatGPT Developer mode, Business/Enterprise rollout notes, Codex CLI, and the API path. - [Docs: Connect Microsoft 365](https://ansvar.eu/docs/setup/microsoft-365): install the published Teams and M365 Copilot agent and verify a cited answer. - [Docs: Connect Gemini](https://ansvar.eu/docs/setup/gemini): Gemini CLI and Code Assist via OAuth, Antigravity, Gemini Enterprise, and the headless API path. - [Docs: Run a TARA](https://ansvar.eu/docs/guides/tara): choose a variant, review the system and risk criteria, assess attack paths, and receive the report. - [Docs: Service credentials](https://ansvar.eu/docs/service-credentials): create, rotate and revoke org-owned OAuth2 client_credentials for n8n, CI jobs and other headless clients. Premium, Team and Company mint them, each with the same gateway tool surface as a signed-in seat on the plan. On Premium each takes a full seat; Team and Company include five per workspace that take no seat. - [Docs: Validate a claim](https://ansvar.eu/docs/guides/validate-claim): claim checking on every plan. `validate_claim` takes the agent's reasoning as a claim graph, refetches every cited provision under the caller's own access, and reports the strength each conclusion can carry, never a verdict. `verify_citations` re-retrieves each reference a report relies on, checks the returned row's identity against it, and compares the quoted text with the served body, one verdict per citation. - [Docs: Tool reference](https://ansvar.eu/docs/reference/tools): the gateway tool surface by family and plan; `describe_capabilities` in a signed-in session returns the live list for that account. - [Coverage JSON](https://ansvar.eu/coverage.json): machine-readable coverage endpoint — the audited jurisdiction, framework, and corpus inventory behind the Coverage page, as JSON. - [MCP discovery](https://ansvar.eu/.well-known/mcp.json): structured site-level pointer to the gateway endpoint, browser-WebMCP tools, and documentation links. - [Plugin manifest](https://ansvar.eu/.well-known/ai-plugin.json): legacy ChatGPT-plugin-format descriptive pointer at the MCP gateway and docs — informational only, no API block. - Browser WebMCP: when a browser supports `navigator.modelContext`, ansvar.eu registers read-only informational tools for agents. Current browser WebMCP tools: - `describe_ansvar`: explains what Ansvar does, the trust model, and important boundaries. - `get_ansvar_gateway_setup`: returns the MCP endpoint, OAuth model, client setup links, and the authenticated-gateway boundary. These tools provide product and setup information. They do not run live gateway queries, execute workflows, upload files, or bypass subscription gates. Live gateway work requires an authenticated Ansvar account through `https://gateway.ansvar.eu/mcp`. ## Company - Contact: team@ansvar.eu - Registered: Ansvar Systems AB, Sweden. - [Partners](https://ansvar.eu/partners): the companies and associations Ansvar works with. Aetheria Partners (reseller and infrastructure partner) and Netsecurity (security services partner) deliver Ansvar to their customers; Ansvar cooperates with the Cloud Security Alliance Romania Chapter. ## Legal - [Privacy](https://ansvar.eu/privacy) - [Terms](https://ansvar.eu/terms) - [DPA](https://ansvar.eu/dpa) - [Subprocessors](https://ansvar.eu/subprocessors) - [ISO Standards Add-on terms](https://ansvar.eu/standards-addon-terms) ## Blog - [Blog index](https://ansvar.eu/blog): field notes on evidence, governed assessments, and defensible decisions. RSS at https://ansvar.eu/blog/feed.xml. - [Open-source steward readiness before the CRA reporting date](https://ansvar.eu/blog/open-source-steward-readiness-cra): A public-source review can help an open-source organisation assess CRA steward status, conditional reporting duties, and the evidence it still needs. - [Our security questionnaire is public: Ansvar Gateway joins the CSA STAR Registry](https://ansvar.eu/blog/csa-star-registry-caiq-transparency): Ansvar Gateway is listed in the CSA STAR Registry — a Level 1 CAIQ v4.0.3 self-assessment with all 261 answers published, including the No answers. How to read it. - [Whine, cringe, protect: emotional feedback channels for AI agents](https://ansvar.eu/blog/agent-affect-channels-whine-cringe-protect): Every team has a complainer — and they're worth having. We gave our AI agents the same license, with three skills you can now use yourself. - [MCP protocol v2: what changes, and how we prepared](https://ansvar.eu/blog/mcp-protocol-v2-what-changes): The Model Context Protocol gets its largest revision on 28 July 2026 — stateless core, an extensions framework, tighter auth. Nothing breaks that day. Here is what changes and what we did ahead of it. - [Citing your own documents: chat-with-your-PDF vs. evidence](https://ansvar.eu/blog/citing-your-own-documents-evidence-not-chat): Every AI tool answers questions about your documents. Almost none can prove, months later, what the cited paragraph said. Why hash-anchored paragraph citations matter for counsel, DPOs, GRC teams, and procurement. - [The citation contract: how article-level citations are validated](https://ansvar.eu/blog/how-article-level-citations-are-validated): A citation that survives audit is a provision URI that resolves, text that matches, and a refusal when neither holds. Inside the cite-resolve-compare pipeline. - [Choosing an MCP client for a compliance team: Claude, Copilot, Cursor](https://ansvar.eu/blog/choosing-an-mcp-client-for-compliance-teams): Claude Desktop, VS Code Copilot, Copilot Studio, and Cursor compared for compliance teams: OAuth MCP support, enterprise controls, data-training policies. - [Stop letting threat models die in a wiki: make STRIDE output double as compliance evidence](https://ansvar.eu/blog/threat-model-to-compliance-evidence): A threat model and a NIS2/DORA/ISO gap analysis describe the same system. Map each threat to the measure it satisfies, with article-level citations. - [EU AI Act: your obligations depend on your role, not your tech stack](https://ansvar.eu/blog/eu-ai-act-obligations-by-role): Provider, deployer, importer, distributor, GPAI provider — each EU AI Act role carries its own obligations. The role test, article references, and live dates. - [Why RAG over a document dump fails regulated work](https://ansvar.eu/blog/cited-answers-vs-rag-for-regulated-work): RAG citations are decorative — no provenance contract links chunk to answer. Regulated work needs typed corpus tools, deterministic validation, and refusal. - [Swedish law as an MCP server: how SFS statutes become a queryable corpus](https://ansvar.eu/blog/swedish-law-as-an-mcp-server): 6,041 Swedish statutes from Riksdagen, segmented to section level and served as an MCP — query by SFS number, chapter, and paragraf, every result cited. - [What you can and can't automate in a DPIA](https://ansvar.eu/blog/automating-the-dpia-evidence-trail): A GDPR Article 35 DPIA has an automatable core and a judgment core. AI assembles the cited evidence trail; the DPO signs necessity and proportionality. - [Working through DORA Article 28: third-party obligations, the contract checklist, and what the auditor asks for](https://ansvar.eu/blog/dora-article-28-third-party-mapping): DORA Article 28 sets the third-party obligations; the contract clauses live in Article 30. Each subsection mapped to ISO 27001 and SCF controls. - [NIS2 vs ISO 27001: a clause-by-clause working mapping (and where ISO stops short)](https://ansvar.eu/blog/nis2-vs-iso-27001-clause-mapping): Every NIS2 Article 21(2) measure mapped to ISO 27001:2022 Annex A — and the three real gaps: reporting clock, management liability, supply chain depth. - [What an MCP gateway is, and why compliance work needs one](https://ansvar.eu/blog/what-is-an-mcp-gateway-for-compliance): Chat and RAG hallucinate regulatory citations. An MCP gateway adds routing, fan-out, tier auth, and deterministic citation validation — and when you need one. - [Effective risk: turning the LLM-era CVE firehose into a triage queue you can actually work](https://ansvar.eu/blog/effective-risk-cve-noise-llm-era): AI tooling files more CVEs than any analyst can read. Effective-risk rescoring deterministically scores CVE × asset × controls, citing every score change. - [How we threat-model AI systems: STRIDE meets MCP](https://ansvar.eu/blog/how-we-threat-model-ai-with-mcp): STRIDE was built for 1999 monoliths. How we adapted it for agentic systems and shipped it as a workflow your own AI client runs through the Ansvar gateway. ## Changelog - [Changelog](https://ansvar.eu/changelog): dated product changes — new corpora, gateway capabilities, account and pricing changes. RSS at https://ansvar.eu/changelog/feed.xml; one feed per category at https://ansvar.eu/changelog/feed/.xml. - [Search refuses an over-long query and tells your agent how to split it](https://ansvar.eu/changelog#2026-10-02-search-query-length-limit): search and search_guidance now refuse a query longer than 500 characters before any source is searched. The refusal is a typed error that tells the agent to split the matter into short issue queries. Before, a pasted case description failed inside the corpus and read as a source outage. - [Ansvar Compliance Skills is in Anthropic's plugin directory](https://ansvar.eu/changelog#2026-09-28-compliance-skills-plugin-directory): The ansvar-compliance-skills plugin is listed in Anthropic's plugin directory. One install adds our task skills and the Ansvar gateway connector to Claude, Cowork and Claude Code. - [Slovenian law: the core statutes, consolidated and in force](https://ansvar.eu/changelog#2026-09-25-slovenian-statutes-expansion): The Slovenian statutes corpus now carries the acts a company, its lawyers and its compliance team work with every week: the civil and commercial codes, labour, tax, procurement, procedure, data protection and cyber security, each in its current consolidated text from PISRS, the state's legal information system. - [Claim check: validate_claim tests a legal conclusion against the text it cites](https://ansvar.eu/changelog#2026-09-22-validate-claim-claim-check): Your agent hands the gateway its reasoning as a graph: the provisions it cites, the quotes it took, the facts and assumptions it relies on, and the conclusions it draws. The gateway refetches each provision under your own access, compares the quotes with the served text, and reports what each conclusion rests on. It never answers "passed". - [A new ansvar.eu: four products, one connection](https://ansvar.eu/changelog#2026-09-22-new-site-four-products): The site now explains Ansvar as four products on one connection: legal and regulatory evidence, living security architecture, compliance assessments, and ISMS as Code. Deploy is a first-class choice, hosted or inside your perimeter. Pricing shows how products and plans fit, and every limit is published. - [Canadian Centre for Cyber Security guidance, now served](https://ansvar.eu/changelog#2026-09-17-canada-cyber-centre-guidance): Canada's national cyber security guidance answers through the gateway in English and French: the ITSP, ITSM and ITSAP series, the ransomware playbook and the threat bulletins, each row licensed under the Open Government Licence – Canada and cited to the Cyber Centre's own page. - [Inflected-form search on national law](https://ansvar.eu/changelog#2026-09-13-inflected-form-search-national-law): Search a national law corpus in its own language and a plural, a definite form or a case ending now finds the provision that carries the base word. The legal text, the citations and the content dates did not change; only the search index was rebuilt. - [EIOPA statements, technical-standards reports and joint ESA products join the guidance source](https://ansvar.eu/changelog#2026-09-12-eiopa-lane-widened): The EIOPA lane of the supervisory-guidance source now serves EIOPA statements, technical-standards reports, the older opinions the publisher's listing left untyped, and the Joint ESA products EIOPA hosts. The standalone eiopa-insurance source is retired; its id keeps working as an alias. - [Preparatory works from six more European parliaments, and the closed terms kept whole](https://ansvar.eu/changelog#2026-09-11-european-preparatory-works-six-parliaments): Bills, government drafts and explanatory reports from the Spanish Congreso, the UK Parliament, the Italian Senato, the Czech Poslanecká sněmovna, the Slovak Národná rada and the Lithuanian Seimas join the German, Austrian and Norwegian material, with the closed legislatures served from their own archive. - [EIOPA guidance joins the financial supervisory sources](https://ansvar.eu/changelog#2026-09-11-eiopa-guidance): EIOPA guidelines, opinions, supervisory statements and technical-standard final reports are now searchable beside the EBA and ESMA publications, each with its publisher reference and source link. - [Financial assessment workflows for contracts and readiness](https://ansvar.eu/changelog#2026-09-10-financial-workflow-variants): Prepare DORA ICT contract and register reviews, MiCA CASP readiness assessments, and AMLR readiness assessments with cited evidence and unresolved findings visible. - [Financial supervisory guidance in your agent](https://ansvar.eu/changelog#2026-09-10-esa-guidance): Search available EBA, ESMA and joint-ESA financial supervisory publications through Ansvar, with source references and publication types visible in the results. - [North Rhine-Westphalia's court decisions join the German case law](https://ansvar.eu/changelog#2026-09-09-german-court-decisions-nrw): German court decisions grow from a federal-only corpus to 281,813 decisions across two publishers, with the NRWE database of North Rhine-Westphalia's courts served beside the federal benches. - [The FCA Handbook, now served](https://ansvar.eu/changelog#2026-09-09-fca-handbook): Every sourcebook of the UK Financial Conduct Authority's Handbook is searchable through the gateway at provision level, with a citation to the section on the FCA's own site. - [NIS2 answered from the national law, across the EU](https://ansvar.eu/changelog#2026-09-07-nis2-national-transpositions): Ask about NIS2 in a member state and the gateway now serves that state's transposing statute for governance, risk-management and incident-reporting duties, with the EU text one call away. - [Chinese national law alongside EU regulatory sources](https://ansvar.eu/changelog#2026-09-06-chinese-national-law): Chinese national laws, legislative decisions and State Council administrative regulations are available with Chinese original text and source citations. A new Simplified Chinese page explains how product teams can use them in an EU market-entry review. - [UK cyber security guidance from the NCSC, now served](https://ansvar.eu/changelog#2026-09-05-uk-ncsc-cyber-security-guidance): The National Cyber Security Centre's guidance collections and security advisories join the gateway's cyber security corpus for the United Kingdom, crawled from ncsc.gov.uk under the Open Government Licence and cited page by page, so a UK organisation's NIS, Cyber Essentials or board-level questions answer from the regulator's own text. - [Swedish financial rulebook: Finansinspektionen's FFFS now served](https://ansvar.eu/changelog#2026-09-04-swedish-fffs-finansinspektionen): Finansinspektionen's föreskrifter and allmänna råd (FFFS) are searchable through the gateway beside the Swedish financial statutes, instrument by instrument from FI's own register, so a Swedish bank's compliance question reaches the binding rulebook and not only the acts that delegate to it. - [Dutch disciplinary tribunal rulings, now served](https://ansvar.eu/changelog#2026-09-04-dutch-tuchtrecht-disciplinary-tribunals): Decisions of the Dutch statutory disciplinary tribunals for the medical, legal, notarial, accountancy, bailiff, veterinary and maritime professions join the gateway as full-text, cited rulings. - [Dutch rechtbanken and gerechtshoven join the case-law corpus, with court and date filters](https://ansvar.eu/changelog#2026-09-04-dutch-court-decisions-lower-courts-and-filters): The Dutch court-decisions corpus now carries the district courts and courts of appeal beside the Hoge Raad, the Raad van State, the Centrale Raad van Beroep and the College van Beroep voor het bedrijfsleven, and search can narrow decisions by court and by date.