M
Moonlit

Moonlit is an Amsterdam legal research and legal data company. Its Data Layer gathers legislation, case law and regulator output from 46 jurisdictions, about 56 million documents linked by more than 200 million references and updated daily, with its depth in European Union and Dutch law. It sells that layer three ways. The Moonlit Platform pairs keyword, semantic and hybrid retrieval with Luna, an assistant that answers within the jurisdictions and portals a user picks and cites a source document for every claim.

Workspaces keep a matter's documents, questions and answers in one table, and Monitors answer set questions about each new document from watched sources.

An MCP connector brings the same data into Claude, Copilot, ChatGPT, Gemini and other MCP compatible clients, and a Data API serves legal AI builders including Saga, Legal Mind and Eudia. Every document carries its official source link, in force status, version history and citations in both directions at article level. Buyers named include law firms, in house legal teams, tax advisers, consultants, public bodies, students and legal AI vendors, with Deloitte, KPMG, Houthoff, De Nederlandsche Bank, the Dutch central government and Booking.com among the logos shown.

Moonlit holds ISO/IEC 27001:2022 certification, hosts in the European Union or the United States at the customer's choice, and publishes its prices in euros. Moonlit Legal Technologies B.V. was founded in 2024 as a spin out of Deloitte Netherlands, with investment from Curiosity VC, and Roderick Lucas, its founder, is chief executive.

Capability grades

All 15 axes, graded from public sources on the date shown. Hover a grade to see what the letter means on that axis.

BB on AI CentralityThe models are the engine of a core capability, layered on a product that would still function without them as a document or workflow system.

AI Centrality

How much of the product is actually AI. Whether the machine learning is the mechanism the buyer is paying for or a feature layered onto conventional software, and whether the vendor is specific about which is which.

Luna's cited answers, semantic and hybrid retrieval, and reranking run on language and embedding models, and Monitors use Luna to answer set questions about each new document from watched sources. Workspaces put Luna to work across a table of a matter's documents.

Underneath sits a curated legal database with keyword retrieval, an article level citation graph and in force status, which works without any model and is sold on its own through the Data API.

Source: Vendor Published
BB on Citation Accuracy and Hallucination DisclosureGrounding is real and documented, with linked primary sources and a described retrieval method, short of published accuracy figures an outsider can test.

Citation Accuracy and Hallucination Disclosure

Whether the vendor publishes measured accuracy on citations and assertions, grounds output to primary sources, and says plainly what its system does when it does not know. Legal has a documented public record of fabricated citations reaching filed briefs, so an untested claim of accuracy is not evidence.

Luna answers only from the jurisdictions and portals a user selects, and each claim cites a document that opens in a reader beside its official source, in force status and version. Retrieval is keyword, semantic, hybrid and reranked over a graph of more than 200 million references.

Moonlit says this leaves no room for hallucinations, yet no accuracy rate or test set backs the claim, and its terms warn that output may be incorrect, inaccurate or incomplete.

Source: Vendor Published
CC on Autonomy and Oversight ModelAutonomy is claimed and oversight is asserted without a mechanism. Human in the loop appears as a phrase rather than a described control.

Autonomy and Oversight Model

What the system decides on its own, what a lawyer must approve, and whether the vendor documents where the review point sits. A tool that drafts under review and a tool that files without one are different products and different risks.

Users set the scope of each Luna answer by jurisdiction and portal, can open every cited source in a reader beside the answer, and can add to the Workspace tables that Luna fills in. Monitors answer only the questions a user sets for each new document.

Beyond the terms' statement that output is not legal advice and that the customer answers for its use, no review step, usage limit or route back to a lawyer is described.

Source: Vendor Published
BB on Operational and Outcome EvidenceReal deployment evidence with substance, short of full attribution or measurement: a named customer without figures, or figures without the named customer.

Operational and Outcome Evidence

Named, dated evidence that the product works in production at real firms or legal departments. Case studies with figures and identified customers count. Unattributed testimonials and launch announcements do not.

Saga, which serves more than 500 law firms and legal teams in over 20 countries, grounds its AI in the Moonlit Data API. Legal Mind built a research agent on the API for more than 1,000 Dutch legal professionals, and Eudia announced a partnership.

Moonlit counts more than 8,000 legal tech vendors, governments, lawyers and researchers among its users and shows 30 organization logos, from Deloitte and KPMG to Utrecht, Tilburg and Ghent universities. No customer reports a measured change in research time or accuracy.

Source: Vendor Published
BB on Privilege and Confidentiality PostureSubstantive published commitments on confidentiality and training use, short of the full picture: commonly silence on segregation between users or matters, or on what the underlying model provider may retain.

Privilege and Confidentiality Posture

How client confidences are handled: attorney client privilege and work product treatment, segregation of one client matter from another, whether client data trains any model, and what the vendor commits to in writing rather than in marketing.

The MCP terms bar Moonlit from training its own models on subscriber prompts or questions. They limit usage data to running the service, security and anonymous aggregate monitoring, and request logs leave out question text, while Google and Nebius embed and rerank questions with zero retention and no training.

The platform terms that govern Luna and Workspaces are weaker, excluding questions from confidential information and allowing their anonymous use for platform development and their sharing with AI component suppliers. Privilege, work product and separation between matters are not addressed.

Source: Vendor Published
BB on UPL and Professional Responsibility PostureA real position is published on advice versus tooling, short of full treatment: commonly a disclaimer without the supervision and competence dimension, or silence on jurisdiction limits.

UPL and Professional Responsibility Posture

Whether the vendor is clear that it supplies a tool rather than legal advice, who its audience is, and how it addresses unauthorized practice of law, competence and supervision duties, and jurisdiction limits. ABA Formal Opinion 512 is the reference point. Where the advice line is not the duty a product raises, the axis is read through the nearest professional duty it does raise: judicial conduct rules and the reviewing duty for products sold only to courts, and the duty to bill for time actually spent for products that draft time entries.

Both sets of terms state that output is not legal, tax or regulatory advice and that the customer alone answers for decisions taken on it. Moonlit is sold to law firms, tax advisers, in house teams, public bodies and students.

Nothing addresses a lawyer's supervision or competence duties when relying on Luna, or the limits of answers drawn from jurisdictions outside a user's own.

Source: Vendor Published
DD on AI Governance and Bias DisclosureNo governance position published for a system whose output affects legal outcomes.

AI Governance and Bias Disclosure

Published governance over model behavior: who owns it inside the vendor, what is tested before release, and what is disclosed about disparate output across matter types, parties, or populations.

No AI governance framework, testing regime, accountable owner or finding on uneven output stands behind Luna or the retrieval models. Moonlit's controls cover information security under ISO/IEC 27001, with a CISO on a team of about 19, and a data and AI lead is named among its leadership, but neither is described as owning model behavior.

Source: Operator Verified
BB on AI Safety and Data StewardshipSubstantive published policy covering most of the ground, short of the full set: commonly no named subprocessor list or no stated incident practice.

AI Safety and Data Stewardship

Retention, deletion, access control, and what happens to prompts and documents after they are processed. Whether the vendor states its subprocessors and its incident practice, or leaves the buyer to assume.

For the MCP connector and Data API, retention is set provider by provider. Customer data stays 90 days after a subscription ends and is deleted within a further 90, operational logs are kept at least 30 days and user access logs up to 365, and question text is not stored.

Access runs on role based controls with multifactor authentication for administrators, an independent penetration test runs every year, and the DPA commits to breach notice within 72 hours. The platform terms set no retention period for Luna conversations or Workspaces.

Source: Vendor Published
CC on AI Liability and RecourseLiability is addressed only through a standard limitation clause that disclaims the exposure the product creates.

AI Liability and Recourse

What the vendor stands behind contractually when its output is wrong. Indemnities, caps, carve outs, insurance, and whether any of it is published or only reachable through a negotiated agreement.

The MCP terms limit each party to direct damages, capped at the fees paid in the prior 12 months across all heads of liability, with claims barred after 12 months. The platform terms cap liability at the insurance payout or the prior calendar year's fees, under Dutch law, and make the customer indemnify Moonlit against claims tied to output.

Moonlit disclaims any warranty that AI output will be free from inaccuracies or hallucinations.

Source: Vendor Published
BB on Practice Systems Integration DepthReal integrations exist and are documented, short of depth: named connections without a description of what they actually move.

Practice Systems Integration Depth

How deeply the product reaches into the systems legal work already lives in: document management such as iManage and NetDocuments, Word and Outlook, contract lifecycle management, matter management, e-billing, and court filing systems.

An MCP connector puts the Data Layer into Claude, Claude Code, Copilot, ChatGPT, Gemini and Codex, with quickstart guides and documented tools and parameters. The Data API offers six retrieval endpoints over one schema for every jurisdiction and is built into Saga, Legal Mind and Eudia.

Enterprise adds single sign on and user management. No document management, matter management or Word integration is offered.

Source: Vendor Published
AA on Deployment Model and Data ResidencyDeployment options and data residency are published, including the regions available, what changes between tiers, and where processing happens as distinct from where data is stored.

Deployment Model and Data Residency

Where the software runs and where the data sits. Multi tenant cloud, single tenant, private deployment, on premises, and whether region of residence is a published option or an enterprise conversation.

Customers choose European Union or United States hosting. The European deployment runs in Microsoft Azure West Europe in the Netherlands, where questions, documents, account records and logs are processed and stored.

Embedding runs in Google Cloud's Netherlands region and with Nebius in the European Union, reranking in a European multiregion setting, and the United States deployment keeps its own subprocessor list. Tenancy is not described.

Source: Vendor Published
AA on Security Certifications and Trust CenterCurrent independent attestation with named scope, reachable without a sales call: a trust center carrying reports, dates and the standards actually covered.

Security Certifications and Trust Center

Independent attestation a buyer can pull without a sales call: SOC 2, ISO 27001, penetration test summaries, a trust center with current reports and named scope rather than a badge image.

ISO/IEC 27001:2022 certificate 202506-107, issued by AssuranceLab on June 30, 2025 and valid to June 30, 2028, covers development, operation and support of the legal research platform. The certificate sits on the trust center, the audit report is available on request, and an independent penetration test runs every year.

No SOC 2 report is offered.

Source: Vendor Published
BB on Model Supply Chain DisclosureThe supply chain is partly disclosed: providers named without change notification, or architecture described without the providers.

Model Supply Chain Disclosure

Which models sit underneath, whose they are, where they run, and whether the vendor commits to telling customers when that changes. A legal buyer inherits every dependency it cannot see.

Question embedding runs on a Gemini embedding model through Google Vertex AI and a Qwen3 embedding model through Nebius, and Google's ranking service reranks passages, each with zero data retention and no training. Moonlit gives 30 days' notice of any new or replaced subprocessor. The language model that writes Luna's answers is not named.

Source: Vendor Published
BB on Commercial TransparencyReal pricing is published for part of the range, with enterprise tiers withheld, or the unit and structure are stated without the figure.

Commercial Transparency

Whether a buyer can learn what this costs without entering a sales process: published rates, the unit being charged, what sits behind an enterprise tier, and what implementation adds.

Prices are set in euros, excluding VAT and billed yearly. The MCP connector costs €35.10 a month for 500 calls or €71.10 for 1,200, and the Premium platform plan costs €112.50 a month with unlimited Luna chat and 1,000 Workspace and Monitor answers each. A Student plan costs €6.26, and a free plan gives 100 calls and five Luna chats a week.

Enterprise is priced per team from ten accounts, Data API pricing comes by quote, and fees are indexed each January.

Source: Vendor Published
BB on Firm and Practice CoverageSegment and practice coverage is described with substance, short of the boundaries: what is supported is clear, what is not is left open.

Firm and Practice Coverage

Who the product is actually built for. AmLaw, midlaw, small firm and solo, in house departments, government and courts, and which practice areas are supported rather than merely claimed.

Coverage runs to 46 jurisdictions of legislation, case law, regulator decisions, official gazettes and parliamentary papers, deepest in European Union and Dutch law, across all fields of law. Named buyers are legal AI vendors, law firms such as Houthoff, in house teams at companies such as Booking.com, Akamai and Sonos, tax advisers, consultants, public bodies including De Nederlandsche Bank, and universities and students.

Gaps by jurisdiction or source are not set out.

Source: Vendor Published
Sources on file

3 public documents

The public pages on file for Moonlit, with the recorded signals each one supports and the date it was last read. Open any of them and check the reading against the record.

Pricing

From €35.10/month, billed yearly, excl. VAT

EUR, as published, never converted
  • Moonlit's MCP connector costs from €35.10 a month for 500 calls, billed yearly.
  • MCP Pro costs €71.10 a month, and the Premium platform plan with Luna, Workspaces and Monitors costs €112.50.
  • Students pay €6.26 a month, and a free plan gives 100 calls and five Luna chats a week.
  • Prices exclude VAT, and Enterprise and Data API pricing come by quote.

Plans are priced per account per month in euros, excluding VAT and billed once a year. The entry MCP plan costs €35.10 for 500 calls a month and MCP Pro €71.10 for 1,200 calls, both with source tracing across the full Data Layer. On the platform, Premium costs €112.50 a month with unlimited Luna chat, Workspaces and Monitors at 1,000 answers a month each, and MCP Pro can be added for €35.10. A Student plan costs €6.26 a month with a student or university email.

The Free plan includes 100 MCP calls a month and five Luna chats a week. Enterprise is priced per team from ten accounts, with SSO, user management, an admin dashboard and Data API access on request.

Implementation: Not stated. Enterprise includes priority support and onboarding.

Note: The plan comparison table shows higher monthly prices for the same plans, €6.95, €39, €79 and €125, without explaining the difference. Yearly totals include 21 percent VAT. Fees are indexed each January 1 by up to the services price index plus 2 percent.

Legal Signals

What each signal means

A signal records what public sources say on the date shown. It is not a grade and it is not a recommendation. Where a signal reads Not addressed, it means the index did not locate the material in public sources on that date, which is a statement about disclosure rather than about the product.

Confidentiality and Privilege

Client Data in Training

Can material a lawyer puts into this product be used to train a model?

Never, in the contract

The published terms prohibit training on customer content. Not a policy page, the agreement.

Clause 11.2 of the MCP terms, updated October 28, 2026, rules out training Moonlit's own models on subscriber prompts or questions, and its embedding providers work with zero retention and no training. The platform terms for Luna do not name training but allow anonymous use of questions for platform development.

Source: Vendor PublishedMoonlit will not use Subscriber prompts or queries to train its own modelsAs of Nov 18, 2026Evidence

Prompt and Output Retention

How long does the product keep what a lawyer typed, and can that be set to zero?

Disclosed fixed window

A specific retention period is published and the customer cannot change it.

The MCP connector and Data API keep no question text, hold customer data 90 days after a subscription ends and delete it within a further 90, and keep operational logs at least 30 days. The platform terms give no period for Luna conversations or Workspaces.

Source: Vendor PublishedCustomer data is held for 90 days after a subscription endsAs of Nov 18, 2026Evidence

Ethical Walls and Matter Segregation

Does retrieval respect the firm’s ethical walls, or can the model read across them?

Not addressed

No located public material addresses walls or matter level segregation.

Workspaces can be shared within a team, and Enterprise adds team workspaces, single sign on, user management and an admin dashboard. No restriction by matter or client is described.

Source: Operator VerifiedAs of Nov 18, 2026

Third Party Request and Subpoena Notice

If someone subpoenas the vendor for a firm’s data, does the firm hear about it first?

Notice committed

Terms commit to notice where lawfully permitted. No transparency report located.

The MCP terms allow disclosure of confidential information, including usage data, where law requires it, with prompt notice where possible. The platform terms promise reasonable efforts at prior notice but exclude questions from confidential information. No transparency report is published.

Source: Vendor Publishedwhere required by law (with prompt notice where possible)As of Nov 18, 2026Evidence
Accuracy and Authority

Primary Law Corpus Provenance

Where does the law in this product come from, and does the vendor have the right to use it?

Sources named, basis unstated

Sources are identified without stating the license or rights basis.

Sources are named by portal, among them Rechtspraak.nl, Overheid.nl, EUR-Lex, Légifrance, Gesetze im Internet and the Council of Europe, and the DPA describes the Data Layer as publicly available European legal sources updated daily. A rights basis is not given for each source.

Source: Vendor PublishedAs of Nov 18, 2026

Good Law Verification

Does the product tell you when the authority it just cited has been overruled?

Own treatment signal

The vendor computes and surfaces subsequent history itself, with the method described.

Moonlit computes an in force status, version history and citations in both directions at article level for each document, and Luna confirms in force dates while writing an answer. Treatment of case law, such as reversal or criticism, is not shown.

Source: Vendor PublishedAs of Nov 18, 2026

Refusal and Uncertainty Behavior

What does the product do when the answer is not in the corpus?

Not addressed

No located public material addresses what the product does when it cannot ground an answer.

No behavior is described for a question the Data Layer cannot answer within the jurisdictions a user selects, and no confidence indicator is shown beside an answer.

Source: Operator VerifiedAs of Nov 18, 2026

Fabricated Citation Record

Does a public court record exist addressing fabricated or hallucinated legal citations in output from this product?

None located

No court order, opinion or disciplinary record addressing fabricated or hallucinated legal citations produced by this product has been located as of the date shown. This is a statement about the public record on that one subject, not a finding about the product, and this signal is not a litigation history.

No court decision in the AI Hallucination Cases database maintained by Damien Charlotin names Moonlit or Luna as the tool behind a filing.

Source: Operator VerifiedAs of Nov 18, 2026Evidence
Professional Responsibility

Bar Guidance Alignment

Has the vendor engaged in public with the ethics opinions its buyers are bound by?

Not addressed

No located public material engages with bar or ethics guidance.

No professional conduct opinion on AI use, Dutch, European or other, is cited for Luna or the Data Layer. The terms rest on the statement that output is not legal, tax or regulatory advice.

Source: Operator VerifiedAs of Nov 18, 2026

Billing and Fee Posture

Does the vendor address what happens to the bill when the work takes an hour instead of six?

Not addressed

The product sits inside a lawyer to client fee relationship and no located public material addresses billing, fee or disclosure treatment, with no savings claim published either.

Law firms and tax advisers who bill clients use Luna for research, and nothing addresses how that research is billed to clients or disclosed on an invoice.

Source: Operator VerifiedAs of Nov 18, 2026

Outside Counsel Guideline Readiness

Can a firm get this vendor through a client’s AI clause without a bespoke negotiation?

Subprocessors listed

A current subprocessor or model provider list is published.

A current subprocessor list for the European deployment gives each provider, region, transfer basis and retention period, including Google Vertex AI and Nebius for embedding, and a separate list covers the United States deployment. The DPA is published in the documentation.

Source: Vendor PublishedAs of Nov 18, 2026Evidence

Court Disclosure Support

If a judge’s standing order requires an AI disclosure, can the product produce one?

Partial record

Some elements of the record are available, short of a document level export.

Each Luna answer cites the documents it used, and Workspaces keep a matter's documents, questions and answers together in one table. No export recording the model used or human verification is described.

Source: Vendor PublishedAs of Nov 18, 2026
Contact

Correct a record, or ask how something was graded

Every grade and every signal on this index is drawn from public sources and dated. If a record is wrong, out of date, or missing an artifact the index did not locate, send the source and it will be reviewed and the record redated. Vendors are welcome to submit documentation. Nothing on this index is for sale, including a listing, a placement, or a grade.

AI Legal Index

The AI Legal Index is an independent index that tracks changes to AI vendors in legal. It holds 535 vendors across 9 categories, each graded on the same 15 capability axes and recorded against 12 legal signals, from public sources. No vendor pays for inclusion, placement, or rating.

Index Status
Last index update
November 19, 2026
The AI Legal Index is an editorial reference. It is not a regulatory body, not a law firm, and nothing published here is legal advice or a recommendation to retain or avoid a vendor. Records are verified against published sources, bar guidance and public court records. Where a record reads not addressed, the material was not located in public sources on the date shown. See the Methodology page for evaluation standards and limitations.
© 2026 AI Legal Index
3801 N Capital of Texas Hwy, Ste E240 · Austin, TX 78746