Art Loss Register: Intelligence Source Guide

The Art Loss Register is a private London company operating the largest commercial database of stolen, missing and disputed art, with around 700,000 items registered and roughly 450,000 market checks run each year. It is not a public register: you cannot browse it, you buy searches of it, and a clean result is a service, not a fact about the object.

At a glance

Source	Art Loss Register
Category	Conflict, Crime & Human Security › Cultural Property & Counterfeiting
Homepage	https://www.artloss.com/
Format	`HTML`
Access	Licensed — paid or institutional access The platform catalogue records this as open. That is wrong, and the correction is explained under Access, licensing and what you may do with it below.
Disciplines	Corporate Intelligence, Legal Intelligence
Mission domains	Art & Antiquities Trafficking

Private stolen/looted art registry. — as catalogued in the platform’s own source registry.

The Art Loss Register is operated by The International Art and Antique Loss Register Limited, a company registered in England and Wales and based in London, founded in 1990. It maintains a proprietary database of art, antiques and collectables reported as stolen, lost, looted, subject to title disputes, pledged as collateral, subject to freezing orders or liens, or simply recorded as being in a collection so that the owner is alerted if it appears on the market. Its own published figures describe over 700,000 items on the database and more than 400,000 checks of items on the market each year, run for subscribing auction houses and art fairs as well as law enforcement, museums, dealers, banks, lenders, lawyers, pawnbrokers and private collectors. Two distinct products come out of it. The first is a database check, applied at volume to auction catalogue lots and art fair stands above a value threshold, producing no per-item document. The second is an ALR Certificate for a single object, which involves the internal database check plus checks of external databases and archives, including several relating to the 1933-45 period, plus research by the company's own provenance team. Alongside the main database it operates the Cultural Heritage At Risk Database for objects recorded in situ, and The Watch Register as a specialist service for lost and stolen watches. There is no public API and no public search interface.

The analytical job the ALR performs is not identification of stolen objects. It is the manufacture of due diligence. In most jurisdictions a purchaser's good faith, and therefore their title, depends on whether they took reasonable steps to check an object's status, and the ALR exists to make those steps documentable, dated and purchasable. That is why the certificate matters commercially and why the search fee is the price of a defence rather than the price of information. For an investigator this reframes the source entirely. A negative ALR result tells you that the object was not matched against one private company's holdings on a particular date, and that is a fact about the market's compliance behaviour rather than about the object's history. A positive result, by contrast, is highly informative: it means someone registered a loss, in enough detail to match, and is still looking. For CORPINT work on art-market counterparties, the more valuable question is often not whether an object is on the register but whether the counterparty ever searched, whether the auction house is a subscriber, and whether the certificate produced covers the period the provenance gap falls in. Those are answerable, and they are the questions that decide LEGINT outcomes in title disputes.

Who publishes it, and why that matters

This is a commercial company with a published fee schedule, and the schedule is the most important document about it. Searches cost a per-item fee with annual subscription tiers for low-volume users, auction houses pay a negotiated per-lot rate above a value threshold with a minimum annual fee, and urgent searches carry a substantial premium. Registration of a stolen item costs a small administrative fee, after which the company charges a location fee of five per cent of the registrant's ultimate net benefit if the item is found, or a recovery fee of twenty per cent if the registrant appoints the ALR to act for them in recovering it. Work for law enforcement and for states seeking repatriation is stated to be free of charge. Read that structure carefully, because it defines both the company's reach and its incentives: the register is populated partly because registration is cheap, the market checks are viable because subscribers pay per lot, and the company's revenue on a successful outcome scales with the value of the object recovered. The same organisation therefore holds the database, runs the search, informs the claimant, and may act as the commissioned agent negotiating the settlement. That combination has attracted sustained criticism from art-market lawyers and academics over the years, and whatever view you take of it, an investigator should treat the ALR as an interested party in any matter where it is engaged rather than as a neutral registry.

Provenance is the first question to ask of any dataset and the one most often skipped. Who collects it, what their incentive is, whether they publish a methodology, and whether they correct the record when they get something wrong all bear directly on how much weight a finding drawn from it can carry.

What a record actually contains

The fields you will be working with, what each one means, and whether it is something you can pivot on. Read the meanings carefully — more analysis is wrecked by misreading a field than by failing to find one, and a field that looks like an observation is often an inference.

Field	Type	What it means	Pivot value
`object_type`	enum	The category of item: paintings, drawings, sculpture, antiquities, furniture, silver, clocks, ceramics, religious items, jewellery, watches, arms and armour, tapestries, musical instruments, coins, medals and collectables. Category drives how searchable an item actually is.	Specialist registries for the category; auction house department; customs tariff classification.
`artist_or_maker`	string	Attributed artist, maker or workshop. For named artists this is the strongest matching key in the system; for anonymous, archaeological and ethnographic material it is empty, which is why those categories match so poorly.	Catalogue raisonne, auction records, museum collection databases.
`title_and_description`	string	Free-text identification of the object. Descriptions written by a distressed owner or a police officer at the time of loss are frequently thinner and differently worded than the catalogue entry a dealer later writes for the same object.	Auction catalogue text; provenance literature; police property records.
`medium_and_dimensions`	string	Material, technique and measurements. Dimensions are the most reliable non-image discriminator in art matching and the field most often recorded inconsistently, since framed and unframed measurements differ and units vary.	Physical comparison against an offered object; condition reports.
`image`	array	Photographs of the object supplied at registration. An item registered without an image is very substantially harder to match, and for many older thefts no photograph exists at all.	Visual comparison; image-based search in other registries; publication and exhibition records.
`loss_type`	enum	Whether the entry records a theft, a loss, a wartime or colonial-era looting claim, a title dispute, a lien or collateral pledge, an authenticity concern reported by police or experts, or a positive registration of an object in a collection. This is the field that determines what a match actually means.	Litigation records; insurance claims; restitution claim registers.
`loss_date_and_place`	timestamp	When and where the object was lost or taken. For 1933-45 claims this is often a period and a country rather than a date and an address, and for archaeological material the loss date may precede any record of the object existing.	Historical archives; conflict and occupation chronologies; export licence records.
`police_reference`	string	The crime or incident reference for the reporting force, where a theft was reported. Its presence distinguishes an entry backed by a criminal complaint from one backed only by an assertion of ownership.	National police property databases; INTERPOL national central bureau; prosecution files.
`claimant`	string	The registrant asserting an interest: owner, heir, insurer who has paid out and taken subrogated title, museum, state or lender. Insurers hold title to a great many registered objects, which changes who must be dealt with on a recovery.	Corporate registries; insurance market records; probate and estate records.
`status`	enum	Whether the item remains outstanding, has been located, has been recovered or the claim has been resolved. Entries are stated to remain on the database until recovery or resolution, so an old unresolved entry stays live indefinitely.	Recovery and restitution reporting; case law.
`certificate_reference`	string	The identifier of an issued ALR Certificate for a searched object, with its date. The certificate documents a search on a date against the holdings as they stood then, which is the entire evidential content of the document.	Provenance file; sale contract warranties; insurance underwriting file.
`search_date`	timestamp	When the check was performed. Because registrations are added continuously, a certificate is a statement about a moment, and the gap between search date and transaction date is a real exposure.	Transaction chronology; subsequent re-search before completion.

Coverage — and what is not in it

Coverage is deep in exactly the places the art market and the insurance market operate, and thin everywhere else. Objects with a named artist, a photograph and a Western owner who was insured are well represented, because insurers register subrogated claims routinely and because the market's own compliance processes feed registrations back. Coverage of Nazi-era losses is substantial, supported by the company's provenance research team and by checks against external 1933-45 archives during certificate work. Coverage of archaeological material is structurally poor for a reason that has nothing to do with the ALR: an object looted from an unexcavated site was never recorded, so nobody can report it stolen, and the Cultural Heritage At Risk Database exists as an attempt to close exactly that gap by registering objects in situ before they are taken. Geographic coverage tracks the market, so losses in Europe and North America are far better represented than losses in source countries across Africa, Asia and Latin America, notwithstanding the company's stated free-of-charge work for states seeking repatriation. Update rhythm is continuous: registrations and searches happen daily, which is why a date is attached to every result and why a search a month old is not the same as a search today.

Known blind spots

Absence of evidence here is not evidence of absence. These are the conditions under which Art Loss Register will not show you something that is nevertheless real:

An object nobody knew existed cannot be reported stolen. Illicitly excavated antiquities, undocumented ethnographic material and objects from private collections that were never inventoried are structurally absent, and these are precisely the categories where trafficking is most active.
Losses that were never reported to anyone are absent. Thefts from institutions that did not want publicity, losses in conflict zones where no functioning police force existed, and objects taken during colonial acquisition are underrepresented in a register fed by insurance claims and police reports.
The database is private and unsearchable by you. There is no public interface, no API and no way to establish what is on it except by paying for a search, so you cannot use it to explore, only to test a specific object you can already describe.
Matching depends on description quality on both sides. A loss recorded in a few words by a distressed owner will not match a richly catalogued lot describing the same object, and unattributed, anonymous and mass-produced items are effectively unmatched.
A clean certificate covers this company's holdings and the additional archives it chose to check, on the date of the search. It is not a title search, not a customs check, not an export licence check, and not a statement that the object is legally saleable in the buyer's jurisdiction.
Objects below the market value threshold are not checked at all in the volume auction and art fair services, so a large share of the physical trade in lower-value antiques and antiquities passes through subscribing venues unexamined.
The register records assertions of interest, not adjudicated rights. A title dispute entry means someone claims an interest; it does not mean the claim is good, and treating a disputed-item flag as proof of bad title is as wrong as treating a clean result as proof of good title.
Coverage of the online and social-media secondary market is limited. The checking model is built around catalogued sales at identifiable venues, and objects moving through classified sites, messaging apps and small dealers largely escape it.
Recovery outcomes are shaped by commercial terms. Where a registrant declines to pay a location or recovery fee, or a holder refuses to engage, an object can be located and remain in place, so an entry marked outstanding does not mean the object has not been found.

Write the blind spot into the product. A statement that something “was not observed in Art Loss Register” is defensible; a statement that it “did not happen” is not, and the difference is what survives cross-examination.

Access, licensing and what you may do with it

Access model: Licensed — paid, or institutional access only

There is no free tier and no exploratory access. You submit an item for search through the website account system, supplying the object details and images, and receive a result and, if the checks are satisfied, a certificate. Published fees are a single search at eighty-five pounds, one hundred euros or one hundred and ten dollars plus VAT where applicable, with annual subscriptions at seven hundred pounds for up to ten searches and fifteen hundred pounds for up to twenty-five. An urgent search, answered within one working day where no further provenance research is needed, costs three hundred pounds, or an additional two hundred and fifteen pounds to upgrade a standard search already submitted. Auction houses pay per lot above a threshold typically around fifteen hundred pounds of value, at published regional rates, with a minimum annual subscription. Antiquities, ethnographic, tribal and Asian art carry an additional requirement: the company states a minimum provenance requirement predating 2000 before it will search such items, which is itself a useful signal about where the risk sits. Registering a stolen item costs a small administrative fee, and law enforcement and state repatriation work is stated to be free. Confirm all fees directly, since they are published on a commercial website and change.

Licence

What you buy is a search result and, where issued, a certificate, governed by the company's terms and conditions rather than by any data licence. Two provisions matter to an investigator. First, the terms require the searcher to cooperate in restoring the item or resolving the claim if a match is confirmed, and typically to hold the object pending settlement, which means submitting a search is not a risk-free enquiry: it can create obligations. Second, you receive an answer, not data. There is no right to redistribute, aggregate or republish the register's contents, and building a derived dataset from search results is outside what is granted. For a case file, the certificate is the artefact you retain and it should be stored with the item's provenance record, dated. Read the current terms and conditions before submitting anything on behalf of a client, particularly where the client's interest in the object may conflict with an obligation to hold it.

Rate limits and fair use

Not a technical question here. Throughput is a commercial one: a single search takes days in the ordinary course and one working day at the urgent rate, and there is no bulk interface for a third party to query at volume. Organisations with recurring need should negotiate a subscription rather than buying single searches, and organisations checking large catalogues are the auction and art fair subscribers the per-lot pricing exists for. The etiquette point that matters is different from most sources in this catalogue: because a search can generate obligations and can alert a claimant, do not submit speculative searches on objects you have no legitimate interest in, and do not use the search service as a research tool for someone else's collection.

Licensing changes, and it changes without warning. A dataset that was free for research this year may not be free for commercial or evidential use next year. Confirm the current terms before you build a dependency on it, and record the terms you relied on alongside the data — the licence in force at the time of collection is part of the provenance.

Collecting it

How Art Loss Register is actually pulled, in the order you would set it up. Prefer the bulk or export interface over per-item lookups wherever one exists: it is kinder to the publisher, faster for you, and gives a reproducible snapshot rather than a series of point-in-time answers you cannot reconstruct later.

Method	Format	Cadence	Notes
Single item search	HTML	per transaction	The standard route. Submit object details and images through the account system; expect a result within days, or one working day at the urgent rate where no further research is needed.
ALR Certificate	HTML	per object	The documentary product. Includes external database and archive checks beyond the internal register, and is what belongs in a provenance file and a case record.
Registration of a loss	HTML	on incident	For victims, insurers and institutions. Small administrative fee, with contingent location or recovery fees on a successful outcome, so the commercial terms should be understood before registering.
Subscription checking	bulk	continuous during sale cycles	The volume service for auction houses and art fairs, checking lots above a value threshold. No per-item certificate is issued, which is the distinction most buyers do not realise.
Cultural Heritage At Risk Database	HTML	on inventory	Registration of objects in situ at museums, stores and archaeological sites, so that a subsequently stolen object can be identified. The only route by which unexcavated or at-risk material enters the system before it is taken.

Ingesting it into the platform

Every step below is idempotent and cursor-based: interrupt one and it resumes from where it stopped rather than duplicating rows or losing progress. Collection is recorded per source, so a feed that quietly stops publishing shows up as a stale timestamp instead of silently thinning your coverage.

Register the source as an external check, not a feed — In sources.php, record the ALR as a manual, chargeable check with a per-item cost and a turnaround, rather than as a collectable feed. This is the honest representation and it prevents an analyst expecting automated coverage that does not exist.
Model the object as an entity before you search — Create the object in entity.php with maker, title, medium, dimensions, images and every provenance assertion you hold. The search is only as good as the description you submit, and assembling it properly is where most of the analytical work actually happens.
Record the search as a dated event — Log the submission, the search date and the result through import.php as an event attached to the object, not as a property of it. A clean result has a timestamp and an expiry in practical terms, and storing it as a static attribute loses that.
Attach the certificate to the case — File the issued certificate in cases.php against the object and the transaction, with the fee and the requesting party recorded, so that the due diligence trail is reconstructable years later when a claim surfaces.
Cross-check against the public registries — Run the same object through the accessible sources in parallel and record each result separately, so the file shows which registries were checked and which were not rather than implying a single comprehensive check.
Resolve the counterparties — Use resolve-everything.php and org-profile.php on the consignor, dealer, auction house and any intermediary, and screen them through the sanctions and financial-crime views. In art matters the counterparty is frequently more informative than the object.
Set a re-search trigger — Configure an alert in alerts.php so that any object held or under consideration for longer than a defined period is re-searched before completion, because registrations are added continuously and a certificate ages from the moment it is issued.
Keep positive matches out of automated distribution — A confirmed match creates legal obligations and involves identifiable claimants and holders. Handle it through the case workflow with access controls rather than letting it flow into shared exports or automated reporting.

Registered sources and their last-collected state are listed in sources.php, and the scheduled chain that keeps them current is in automation.php.

How it is wrong, and how to tell

Every dataset is wrong in characteristic ways. Knowing which ways is the difference between using a source and being used by one, and it is the part of source evaluation most often skipped because it is the part that takes work.

Judged as a commercial due diligence instrument, the ALR is the market leader and its results are taken seriously by courts, insurers and auction houses, which is itself a functional measure of quality. Judged as an intelligence source, it is opaque in ways that matter. You cannot inspect the database, you cannot know its completeness for your category or country, you cannot audit a negative result, and the company is both the holder of the data and a commercially interested party in the outcome of any recovery it is engaged on. The positive findings are strong: a match means a specific loss was registered with enough descriptive detail to be matched, usually with a photograph and often with a police reference, and those matches have repeatedly held up. The negatives are weak and should never be described as clearance. The soundest basis for judging any individual result is procedural rather than substantive: what was searched, on what date, against which additional archives, and with what provenance information supplied by the submitter. A certificate issued on thin provenance information supplied by an interested seller is a much weaker document than one issued after the company requested and received supporting evidence, and the certificate itself does not always make that distinction visible.

Characteristic false positives

A clean result is read as clean title. It means no match was found in one private database plus whatever additional archives were checked, on one date. Title depends on facts the register has never seen, including export legality, cultural property law in the source country and the chain of possession.
Description mismatch produces a false negative. The same object described as a school-of attribution by one party and as an autograph work by another, or measured framed rather than unframed, can fail to match a live registration, and the failure is invisible in the result.
A match on a title dispute entry is read as evidence of theft. Disputed items, collateral pledges and authenticity concerns sit in the same database as thefts, and conflating them defames a legitimate holder and misdirects an investigation.
The volume auction check is mistaken for a certificate. Lots checked in a subscribing auction house's catalogue receive a database check only, with no external archive research and no per-item document, and buyers routinely assume more was done than was.
Objects below the value threshold are assumed checked. Volume checking applies above a threshold, so a low-value lot in a subscribing sale may never have been searched at all, and the venue's subscription tells you nothing about that individual lot.
An old registration is treated as current. Entries remain until recovery or resolution, so an object may be listed as outstanding long after it was quietly returned, settled or recovered without the register being updated by the registrant.
The register is treated as comprehensive for antiquities. It is not, and the company's own requirement for pre-2000 provenance before searching such items is an acknowledgement that a database check alone cannot resolve their status.
A search is treated as a confidential enquiry. Submitting an object generates a record, can trigger contact with a claimant on a match, and carries contractual obligations, which makes it a step with consequences rather than a neutral lookup.

None of these make the source unusable. They make it a source that requires corroboration before an assertion built on it goes into a product, which is true of every source and admitted by few.

Ageing

Results age from the moment they are issued, and the ageing is asymmetric. A negative result decays continuously, because registrations are added every day and an object stolen last week can be registered next month; a certificate obtained at the start of a six-month consignment period is materially weaker at the end of it than at the beginning, and re-searching before completion is the standard mitigation. A positive result does not decay in the same way, but the claim behind it can be resolved, settled, withdrawn or superseded by litigation without the register necessarily reflecting it, so a match found in an old file should be verified as still live before it is acted on. The other thing that ages is the description: objects are cleaned, restored, reframed, reattributed and renamed, and an entry recorded thirty years ago in the vocabulary of the time may not resemble the same object as catalogued today. A stale record here looks like an outstanding entry for an object that was quietly returned years ago, or a clean certificate that predates the theft it was supposed to catch.

What this source feeds

A source is only worth what it lets you conclude. These are the disciplines that collect through it, the mission domains it serves and the data points it yields — every one is a tag, so you can follow any thread from here into the rest of the library.

Collected by these intelligence disciplines

Serves these mission domains

Art & Antiquities Trafficking

Yields these data points

Image / Photograph

How each sector uses Art Loss Register

The same dataset is worked very differently depending on who you are, what authority you hold, and what you are ultimately producing. A military analyst is supporting a commander’s decision; a journalist is meeting a publication standard; an NGO caseworker is protecting a person. The records are shared — the constraints, thresholds and outputs are not.

🎖 Military and defence

Relevant to cultural property protection responsibilities in occupied or post-conflict territory, where forces have obligations under the law of armed conflict to prevent looting and to avoid becoming a channel for it. The practical uses are narrow but real: pre-deployment awareness of what is at risk in a theatre, a channel through which recovered material found in the course of operations can be checked and reported, and a supporting record where a state seeks repatriation afterwards. Note that the register's weakest coverage, undocumented archaeological material, is precisely what conflict looting produces, so an unmatched object recovered in theatre carries no implication of clean origin. Recovery, custody and handover should follow the legal advisers and the cultural property protection channels rather than a commercial database result.

🕵 National intelligence

Cultural property trafficking sits alongside other illicit flows and shares its infrastructure: the same freight agents, free ports, shell companies and intermediaries. The ALR is a narrow but occasionally decisive input into that picture, best used to test a specific object and to establish whether a counterparty engages with market compliance at all. The more productive intelligence question usually concerns the network rather than the object, and the register's private, commercial character means it is a supporting check rather than a collection platform. Where the matter touches sanctions or conflict financing, treat the company as an interested private party and corroborate anything material through law enforcement channels and public records.

👮 Law enforcement

For investigators this is a working tool with a specific function: registering losses so that they are checked against the market at scale, and testing seized or offered objects against outstanding claims. The company states that its work for law enforcement is free of charge, which removes the cost barrier but not the evidential one. Anything obtained from a private register is intelligence, not evidence, and a match should be pursued through the reporting force, the national central bureau and INTERPOL's own database rather than through the commercial relationship. Where a recovery is in prospect, be alert to the difference between a claimant's private commercial settlement and a criminal disposal; a negotiated return can extinguish the prosecution's best evidence, and that decision should be taken consciously.

🔍 Private investigation and corporate security

This is the standard instrument for pre-purchase and pre-loan due diligence and, in many art-market engagements, is what the client is actually buying. Do the work properly: assemble the fullest possible description and images before submitting, decide whether an urgent search is justified by the transaction timetable, and re-search before completion if the interval is long. Be candid with the client that a certificate is a defence rather than a guarantee, and that the categories where their risk is highest, antiquities and ethnographic material, are the categories the register covers least well. Where a client's interest may conflict with the obligation to hold a matched object, raise that before the search rather than after.

📰 Journalism and OSINT media

Useful for verification and unusable for exploration, since there is no public interface and no way to browse. In practice the register enters reporting in three ways: an object is confirmed as registered stolen, a sale proceeded without a check, or a certificate turns out to have been issued on inadequate information. All three are legitimate stories and all three require care, because a database match is an assertion by a claimant, not an adjudication, and naming a current holder as in possession of stolen goods before a claim is resolved is a defamation risk in most jurisdictions. Note also that the company is a commercial actor with a stake in recoveries it handles, and that its own account of a case is the account of a participant.

🌍 NGO, humanitarian and human rights

For heritage protection organisations and source-country advocates, the important engagement is not searching but registering. The Cultural Heritage At Risk Database is the mechanism by which objects still in situ, in museums, stores and on sites, can be documented so that theft becomes detectable at all, and the company states this service and repatriation work for states are provided free. That is the highest-leverage activity available, because the structural failure in cultural property protection is that undocumented objects cannot be reported stolen. Alongside it, advocate for Object ID standard documentation and for deposits into the public registries, so that protection does not depend on one private company's continued goodwill or continued existence.

🎓 University and research

Provenance researchers, art historians and criminologists all encounter the ALR, and its private character is itself a research subject: a market that discharges its due diligence obligations through a paid check against an unauditable database is a governance arrangement worth studying. For provenance work the certificate process is more useful than the result, since it involves 1933-45 archive checks that a researcher may replicate independently through the public restitution registers. Do not treat aggregate claims about the register's size or hit rate as findings; they are company-published figures without external audit. And be careful in publication: naming current holders of disputed objects raises both legal and ethical questions that a university's research ethics process should see before the article does.

Playbook: working Art Loss Register end to end

A repeatable sequence from first pull to finished product. Each phase states what you are trying to establish, not merely what to click — the objective is a defensible chain of reasoning, not a completed checklist.

Phase 1 — Decide what question the search is actually answering

A search establishes whether a described object matches this company's registrations on a date. It does not establish title, export legality or authenticity. Writing down which question you are answering, and which three you are not, prevents the certificate being over-read later by a client or a court.

Phase 2 — Build the object description before spending anything

Assemble maker, title, medium, exact dimensions with a stated convention, marks and inscriptions, condition, and the best available photographs from multiple angles. Matching quality is determined entirely at this step, and a thin submission buys a weak negative.

Phase 3 — Reconstruct the provenance chain independently first

Establish who owned the object when, with documents rather than assertions, and identify the gaps. For antiquities the company requires provenance predating 2000 before it will search, so this work is a precondition rather than an afterthought, and the gaps you find will usually be more informative than the search result.

Phase 4 — Check the free and public registers before the paid one

Run the object through INTERPOL's database via its public application, the relevant national police registers, the wartime-losses registries and the specialist category databases. These cost nothing, cover different populations, and frequently answer the question before a fee is incurred.

Phase 5 — Understand what the venue actually did

If the object is offered by a subscribing auction house or at a subscribing fair, establish whether it fell above the checking threshold and whether a certificate was issued or only a database check performed. Buyers routinely assume a certificate exists where none does, and this distinction decides how much weight the sale's compliance carries.

Phase 6 — Consider the consequences before submitting

A search creates a record, may trigger contact with a registered claimant, and under the terms may oblige you to hold the object and cooperate in resolving a claim. Where your client is a current possessor, brief them on that exposure and obtain instructions before submission, not after a match.

Phase 7 — Choose the search product deliberately

A standard search, an urgent search at premium rate, and a full certificate with external archive research are different products at different prices answering different questions. Match the product to the transaction: a same-week completion justifies the urgent fee, a Nazi-era provenance gap justifies the certificate.

Phase 8 — Investigate the counterparty in parallel

While the search runs, profile the consignor, dealer and intermediaries: corporate ownership, sanctions exposure, litigation history, prior involvement in disputed sales. In art matters the counterparty's record predicts problems better than the object's description does, and this work is often decisive when the search comes back clean.

Phase 9 — Interrogate a positive match rather than accepting it

Establish what kind of entry matched: a theft with a police reference, a title dispute, a lien, an authenticity concern. Obtain the registrant's identity where you can, check whether the underlying claim is still live, and verify against the reporting force. A match is the start of an investigation, not its conclusion.

Phase 10 — Handle a live match through the proper channels

Where the entry reflects a reported theft, the route runs through law enforcement and, internationally, through national central bureaux and INTERPOL, not through a private negotiation that suits the commercial interests of the parties. Record who was notified and when, and take advice before any object changes hands.

Phase 11 — Re-search before completion and document the interval

Registrations are continuous, so a certificate is a snapshot. For any transaction where months pass between search and completion, run a fresh search and file both results, with the dates. This is cheap insurance and it is the step that distinguishes a diligent file from a decorative one.

Phase 12 — Write the limits into the report

Whatever you deliver should say which registers were checked, on what dates, what each covers and does not, and that the absence of a match is not evidence of clean title. Without that paragraph the certificate will be quoted back to you as a clearance you never gave.

The platform ships this as a step-checked workflow in playbooks.php, so progress is recorded against a case rather than held in someone’s head.

What to pair it with

No single source carries a finding. These are the datasets that corroborate, extend or contradict this one — and a source that contradicts is worth more than one that agrees, because it is the only thing that will tell you when you are wrong.

Source	Relationship	What it adds
INTERPOL Stolen Works of Art Database	corroborates	The international police register, populated only through national central bureaux and partner bodies, and publicly searchable through the ID-Art app. Different population, no fee, and police-certified.
ICOM Red Lists	extends	Category-level typologies of cultural objects at risk from specific regions. Tells you whether an object belongs to a class that warrants heightened scrutiny even when no individual registration exists.
Central Registry of Information on Looted Cultural Property 1933-1945	extends	The principal public resource for wartime losses and restitution claims, covering the period where the ALR's certificate research is most active and most contested.
German Lost Art Foundation Lost Art Database	corroborates	Public register of cultural property lost, moved or seized in the Nazi era and, increasingly, colonial contexts. Free, searchable and complementary to any 1933-45 provenance gap.
Art Recovery International	contradicts	A competing private registry and recovery business founded by former ALR personnel. Its existence matters analytically: an object absent from one private register may be recorded in another.
The Watch Register	extends	The company's own specialist database for lost and stolen watches, which is the correct instrument for that category rather than the general register.
Carabinieri Command for the Protection of Cultural Heritage	corroborates	Italy's specialist cultural property police maintain one of the largest national stolen art databases, and Italian material is far better covered there than in any commercial register.
UNIDROIT 1995 Convention	prerequisite	The instrument that makes due diligence, including consultation of accessible registers, legally consequential for a possessor's claim to compensation. It is why the certificate has commercial value.
National Stolen Art File	corroborates	The United States federal register of stolen art, maintained by the FBI's Art Crime Team, covering objects reported through US law enforcement.

Legal, ethical and operational constraints

Three distinct legal frames apply and they interact badly. The first is the law of title and good faith purchase, which differs sharply between jurisdictions: in some civil law systems a good faith purchaser can acquire title to stolen property after a period, in most common law systems they cannot, and the 1995 UNIDROIT Convention conditions a possessor's right to compensation on having exercised due diligence including consultation of accessible registers. That is the legal machinery the ALR certificate is designed to engage, and it is why the document is worth more than the information in it. The second frame is contract: the terms of a search bind the searcher, including obligations to cooperate and typically to hold a matched object, so a search is a legally consequential act and should be taken with the client's informed instructions. The third is defamation and privacy, since a register entry is an assertion by a claimant and publishing that a named individual or business holds stolen property before any adjudication is actionable in most jurisdictions. Add to all three the cultural property regimes of the source country and the import controls of the destination, neither of which any register checks.

Operational security

Searching is not anonymous and should not be assumed to be. The company knows which object you asked about, when, and on whose behalf if you told it, and where an object matches a registration the process is designed to put the registrant in contact with the current holder, which necessarily discloses that someone is looking at the object. For an investigator working discreetly, that is a material consideration: a search can alert a claimant, a dealer or, through them, the subject of your enquiry. Where discretion matters, exhaust the public registers and the documentary provenance work first, decide consciously whether the search is worth the disclosure, and consider having it submitted by the party who has a natural commercial reason to be asking. Do not submit searches on objects belonging to a third party you are investigating as a way of probing their holdings; it is a misuse of a due diligence service and it is traceable to you.

Two rules that hold regardless of jurisdiction. Collection that is lawful is not automatically proportionate, and a dataset assembled for one purpose does not carry consent for another. Where the records concern identifiable people, the question is not only whether you may hold the data but whether holding it serves the purpose you are accountable for.

Is it earning its place?

Sources accumulate. Feeds get added during an incident and are never reviewed again, and a decade later the pipeline is carrying dead weight that nobody dares remove. These are the measures that show whether Art Loss Register is contributing anything, and they are worth baselining now so the answer is available later.

Proportion of objects in your matters searched before rather than after the transaction, which is the difference between due diligence and post-hoc justification.
Median interval between search date and transaction completion, and the share of transactions where a fresh search was run before completion.
Share of searches submitted with complete descriptions and multiple images, tracked because match quality is determined by submission quality and nothing else.
Number of matters where a public register produced a finding the paid search did not, which tells you honestly how much the fee is buying in your object categories.
Number of matters where the counterparty investigation, rather than the object search, identified the problem, as a check on where your analytical effort should go.
Cost per matter attributable to register checking, against the number of findings it produced, reviewed annually rather than assumed.
Share of your reports that state explicitly what a clean result does and does not establish, which should be all of them.

Beware of volume. Indicator counts rise easily and say almost nothing. Unique contribution — findings this source produced that no other source in your stack would have — is the measure that matters, and it is usually far lower than anyone expects.

Tradecraft notes

The distinctions that separate a competent analyst from a fast one:

A clean certificate is a purchased defence, not a fact about the object. Read it as evidence that a process was followed, and never let it appear in your report as clearance.
The description you submit determines the answer you get. Time spent on dimensions, marks, condition and photographs before submission does more for the result than any choice between search products.
Ask what kind of entry matched. Theft, title dispute, lien and authenticity concern all live in the same database and mean entirely different things about the object and about the people involved.
Establish whether the venue actually checked this lot. Auction and fair subscriptions apply above a value threshold and produce no per-item document, so a subscribing venue is not the same as a checked object.
The register's blindest spot is the trafficker's best asset. Objects that were never documented cannot be registered as missing, which is exactly why antiquities move through the market with clean search results.
Treat the operator as an interested party. It holds the database, performs the search, contacts the claimant and may act as the paid recovery agent, and that combination should shape how you weigh its account of a case.
A search is an act with consequences. It creates a record, may trigger contact with a claimant, and may oblige you to hold the object; brief the client before submitting rather than explaining afterwards.
Run the free registers first. INTERPOL's public application, national police registers and the wartime-losses databases cost nothing, cover different populations and frequently answer the question outright.
Investigate the people as hard as the object. Consignor, dealer and intermediary records predict problems better than object descriptions do, and they are available to you without a fee.

Questions analysts actually ask

Can I search the database myself for free?

No. There is no public interface, no API and no browsing. You submit a described object and pay a fee, and you receive a result. If you need a register you can actually search yourself, INTERPOL's Stolen Works of Art database is accessible free through its ID-Art application and through an authorised-user account.

What does an ALR Certificate actually prove?

That a described object was checked against this company's database and certain additional archives on a stated date and no match was found or no problem was apparent from the provenance supplied. It is designed to evidence due diligence in a later title dispute. It proves nothing about export legality, cultural property law in the source country, or authenticity.

Is a lot in a subscribing auction house's catalogue checked?

Only if it is above the value threshold that the subscription covers, and even then the volume service is a database check without the additional archive research or the per-item certificate. The company draws this distinction explicitly, and buyers routinely fail to.

What happens if my object matches something on the register?

Under the search terms you are generally required to cooperate in restoring the item or resolving the claim, and will usually be asked to hold it pending settlement. This is why a search should be submitted with the client's informed instructions. Where the entry reflects a reported theft, involve law enforcement rather than negotiating privately.

Why are antiquities treated differently?

Because a looted archaeological object was never recorded before it was taken, so it cannot appear on any stolen-object register. The company requires a minimum provenance predating 2000 before it will search such items, which is a candid acknowledgement that a database check cannot resolve their status. For this category, documentary provenance and the Red List typologies do the real work.

Who registers objects, and does that shape what is on the database?

Owners, insurers who have paid a claim and taken title, museums, states and law enforcement. Insurance subrogation is a major channel, which biases coverage towards insured Western collections and away from losses in places with no insurance market. Positive registrations of objects still in collections are a separate category that also sits on the database.

How much does it cost?

The published schedule at the time of writing is eighty-five pounds, one hundred euros or one hundred and ten dollars plus VAT for a single search, with annual subscription tiers for up to ten and up to twenty-five searches, an urgent search at three hundred pounds, and negotiated per-lot rates for auction houses above a value threshold. Registration of a stolen item carries a small administrative fee plus contingent location or recovery fees. Confirm current pricing directly.

Should law enforcement use it?

Yes, and the company states its law enforcement work is free of charge. Use it to register losses so they are checked against the market at scale, and to test seized objects. Treat any result as intelligence rather than evidence, and pursue matches through the reporting force and the national central bureau so that a criminal disposal is not foreclosed by a private settlement.

Is it the only register worth checking?

No, and treating it as such is the most common failure in art due diligence. INTERPOL's database, national police registers, the wartime-losses registries, category specialists and the ICOM Red Lists all cover different populations, and several are free. A defensible file records what each of them said, on what date.

Standards, formats and interoperability

What this source speaks natively, and what it has to be translated into before a partner can consume it. Work that arrives in a recognised format is easier to defend, easier to hand over and easier to automate against:

Object ID, the international documentation standard for describing cultural objects, is the vocabulary any registration or search description should follow; it is what makes an object identifiable across registers.
The 1995 UNIDROIT Convention on Stolen or Illegally Exported Cultural Objects makes consultation of accessible registers part of the due diligence test that governs a possessor's entitlement to compensation.
The 1970 UNESCO Convention frames the import, export and transfer of ownership rules that a register check does not address and that a buyer still has to satisfy.
National good faith purchase and limitation regimes determine what a due diligence certificate is worth, and they differ enough between jurisdictions that the same certificate has different legal value in different countries.
Auction house and art fair vetting protocols define the value thresholds above which lots are checked, and are the practical mechanism through which most market checking actually happens.
Anti-money laundering regimes covering art market participants in several jurisdictions impose customer due diligence obligations that are distinct from, and not satisfied by, an object search.
Platform exports carry object entities, counterparties and case records in STIX 2.1, MISP, CSV, JSON and JSONL, so art matters share structure with any other investigation.

References

Primary documentation and authoritative references for this source. Publishers revise and retire material, so treat the retrieval date as part of the citation and re-check before relying on any of it in a formal product.

Art Loss Register — The International Art and Antique Loss Register Limited. The company's own site, with its published figures for database size and annual checks. The starting point, read as the marketing of a commercial service rather than as a neutral description.
Art Loss Register FAQs — Art Loss Register. The most informative page on the site: current fee schedule, the distinction between a certificate and a catalogue check, the location and recovery fee structure, and the antiquities provenance requirement.
Search service — Art Loss Register. How a search is submitted and what information is required. Read before assembling an object description, since submission quality determines match quality.
Registration service — Art Loss Register. The route by which losses enter the database, including the categories of claim accepted and the Cultural Heritage At Risk Database for objects in situ.
Recovery and mediation service — Art Loss Register. The commissioned recovery business. Worth reading specifically to understand the incentive structure surrounding any recovery the company is engaged on.
The Watch Register — Art Loss Register. The specialist watch database operated by the same company, and the correct instrument for that category.
INTERPOL Stolen Works of Art Database — INTERPOL. The public, police-certified alternative, with a different population and no fee. Any file that checked only the commercial register is incomplete.
UNIDROIT Convention on Stolen or Illegally Exported Cultural Objects — UNIDROIT. The convention text and status list. The legal basis on which a documented register check affects a possessor's position.
Central Registry of Information on Looted Cultural Property 1933-1945 — Commission for Looted Art in Europe. The principal public resource for wartime losses, and the free counterpart to the archive research bundled into a certificate.
Lost Art Database — German Lost Art Foundation. Public, searchable register of cultural property lost or seized in the Nazi era and, increasingly, colonial contexts.
ICOM Red Lists — International Council of Museums. The category-level risk typologies that tell you when an unmatched object still warrants scrutiny, which is the question a clean search result cannot answer.

Link integrity: every reference above was verified with a live request when this page was generated. Where a publisher had moved or withdrawn a document, the link was repointed at a preserved copy in the Internet Archive and marked as archived. Anything with no reachable copy anywhere had its link removed rather than left to rot — the source is still credited, it simply cannot be linked.

Put it into practice

The Quantus Intel threat intelligence platform operationalises this source: it models the object as an entity with its full provenance chain, records every register check as a dated event rather than a permanent attribute, profiles the counterparties alongside the object, and triggers a fresh check before completion so a due diligence file cannot quietly expire.. Browse the full source catalogue, or follow any tag above into the rest of the library.