About CasinoGuard
An immutable, evidence-backed canonical dataset for the online casino industry. Built for AI. Grounded in verifiable evidence.
CasinoGuard is an independent, AI-first dataset that collects, preserves, and structures factual information about online casinos.
Built for artificial intelligence systems, retrieval-augmented generation (RAG), autonomous agents, search engines, developers, and researchers, CasinoGuard provides structured claims supported by traceable evidence, source provenance, geographic observations, and historical records.
CasinoGuard is not an affiliate website, casino review platform, ranking service, or promotional directory. It does not publish casino recommendations or sell favourable findings.
Its purpose is to establish a canonical record of what has been observed, what the available evidence supports, and what remains unknown.
Evidence is preserved. Claims are evaluated. Nothing is accepted on authority alone.
Why CasinoGuard exists
Information about online casinos is fragmented, frequently changing, and often difficult to verify independently.
A casino may identify one operator on its website, display different licensing information to visitors in different countries, or change its bonus conditions without preserving previous versions.
Regulatory records, casino disclosures, third-party publications, and archived information may present different accounts of the same subject.
Traditional casino websites typically consolidate this information into editorial descriptions, ratings, and recommendations. These formats can obscure the distinction between independently confirmed facts, statements made by casino operators, and information repeated from other publications.
For AI systems, this creates a fundamental problem: a confidently written statement is not necessarily an established fact.
CasinoGuard addresses this problem by maintaining structured, evidence-backed records in which every published finding can be traced to identifiable sources and observations.
The objective is not to produce another interpretation of the casino industry. It is to make the underlying information independently examinable.
A canonical source of casino data
CasinoGuard organises information around casino entities, atomic claims, original observations, and preserved evidence.
Rather than maintaining competing descriptions of the same casino, the dataset provides a consistent structure for representing its observable characteristics.
These include operating companies, licensing arrangements, domains, payment methods, game providers, promotions, withdrawal conditions, supported currencies, geographic availability, and other verifiable information.
Each finding is associated with the evidence used to establish it.
Where sources disagree, the disagreement is recorded. Where information cannot be confirmed, the uncertainty remains explicit.
Canonical does not mean unquestionable or permanently correct. It means that information is organised into a consistent, traceable record with identifiable claims, evidence, and historical context.
CasinoGuard is a canonical record of evidence-backed findings, not an authority that declares facts without proof.
Evidence, not assertions
CasinoGuard operates on a fundamental distinction: what a source says is not necessarily what independent evidence establishes.
A casino may claim to hold a gambling licence. Its website may identify a particular company as its operator. Its promotional pages may advertise specific bonus conditions.
CasinoGuard records these disclosures as source assertions and evaluates them against the available evidence.
A casino's statement about its own licence is not equivalent to confirmation from the relevant licensing authority.
Likewise, observing a payment method in a casino's cashier establishes that it was displayed under the recorded conditions. It does not necessarily prove that a transaction can be completed successfully.
This separation between source assertions, direct observations, and independently confirmed information is central to the dataset.
Every finding is intended to preserve the distinction between:
- What a source reports.
- What CasinoGuard directly observes.
- What independent evidence confirms.
- What remains disputed or unknown.
CasinoGuard does not transform repetition into verification. Multiple pages repeating the same unconfirmed statement do not automatically establish its truth.
Immutable evidence, evolving knowledge
CasinoGuard is built around an immutable historical evidence record.
When source information is collected, the original observation and its associated evidence artifacts are preserved rather than silently replaced by subsequent changes.
Depending on the collection method, evidence may include screenshots, full-page captures, web archives, original API responses, HTTP metadata, source URLs, timestamps, geographic context, and cryptographic file hashes.
Each evidence record is assigned a persistent identifier.
When a casino changes its operator disclosure, licence information, payment methods, or promotional terms, new observations are recorded without erasing the earlier evidence.
Historical evidence and current findings remain distinguishable.
The dataset's interpretation of the available evidence can evolve. Claims may be reassessed, corrected, or superseded when new information becomes available.
Evidence is preserved. Knowledge evolves. History is not rewritten.
Immutability describes the preservation of the historical evidence record. It does not imply that every published finding is permanently accurate or that new evidence cannot change a conclusion.
Atomic, evidence-backed claims
CasinoGuard represents information through atomic claims rather than subjective editorial opinions.
A claim describes a specific, independently examinable finding about a casino.
For example, a claim may identify the operating company named in a regulatory register, the payment methods displayed to visitors from Finland, or the wagering requirements published in a particular bonus offer.
Each claim is associated with relevant evidence, verification status, and contextual information.
Claims are designed to be individually retrieved, examined, compared, and cited without requiring an AI system to interpret an entire casino review.
Supporting and conflicting observations remain traceable.
Where a finding aggregates information from multiple countries or sources, the underlying observations remain identifiable.
This structure enables AI systems to retrieve both the finding and the evidentiary basis on which it rests.
Independent of commercial influence
CasinoGuard is independent of casino operators, affiliate networks, and promotional ranking systems.
The dataset does not sell rankings, favourable verification outcomes, sponsored recommendations, or preferential claim treatment.
Casino operators, advertisers, and other interested parties have no entitlement to change findings, suppress unfavourable observations, or override evidence-based verification rules.
Published findings are intended to be determined by the available evidence and the dataset's documented methodology, not by commercial interests.
The same verification principles apply regardless of a casino's ownership, reputation, market position, or commercial relationships.
A source may provide new evidence or identify an error. Any resulting correction must be justified by the evidence, not by the identity or influence of the party requesting it.
No commercial relationship establishes truth. No payment substitutes for evidence.
Automated collection, transparent provenance
CasinoGuard uses automated workflows to collect, preserve, and evaluate information from casino websites, regulatory sources, official documents, and other relevant materials.
These workflows record the context in which information was obtained, including the source, collection time, geographic location, and available technical metadata.
The objective is to reduce subjective editorial intervention and make the process traceable.
Automated collection does not eliminate the possibility of technical errors, misleading source information, or incorrect interpretation.
CasinoGuard therefore treats source provenance, conflicting evidence, and verification limitations as essential parts of the published data.
A finding is not considered reliable merely because an automated system produced it.
Its reliability depends on what the evidence actually establishes.
Geographic and historical context
Online casino information is not always globally consistent.
A casino may display different operators, licensing disclosures, payment methods, promotions, or terms to visitors in different countries.
CasinoGuard records observations within their geographic context rather than assuming that information collected from one location applies everywhere.
Country-specific findings remain distinguishable from aggregated findings.
An aggregated record combines the observations available in the dataset. It does not imply that every country has been examined.
Time is treated with the same precision.
Information observed on a particular date is not automatically presented as permanently current. Historical observations remain identifiable, and older evidence can be flagged as potentially outdated.
This enables researchers and AI systems to distinguish between current findings, historical information, and geographic differences.
Built for AI retrieval and citation
CasinoGuard is designed as knowledge infrastructure rather than a conventional content publication.
Its architecture supports the retrieval of individual findings together with their supporting evidence and relevant limitations.
The dataset provides:
- Canonical casino records — consistently structured information about individual casinos.
- Atomic claims — independently retrievable findings with traceable evidentiary support.
- Persistent identifiers — stable
ev_references for individual evidence records. - Source provenance — identifiable origins, collection timestamps, and preserved artifacts.
- Explicit verification — clear distinctions between verified, supported, reported, disputed, and unknown information.
- Geographic observations — country-specific findings and transparently scoped aggregations.
- Historical records — preserved observations and traceable changes over time.
- Machine-readable access — structured JSON, Markdown, and API representations.
CasinoGuard is intended for use in retrieval-augmented generation systems, autonomous AI agents, AI-powered search, research workflows, and structured data applications.
Each casino record contains consistently organised sections with self-contained factual summaries and references to supporting evidence.
The objective is to make source-grounded information easier for AI systems to retrieve, interpret, and cite.
CasinoGuard cannot guarantee that a downstream AI model will never produce an incorrect answer. It provides the evidence and provenance needed to make factual claims more independently verifiable.
Open evidence and traceability
CasinoGuard makes the evidentiary basis of its published findings accessible for independent examination.
Evidence records are identified by persistent ev_ identifiers, the stable references to cite. A cl_ identifier is not a permanent record: it is generated from the claims behind a statement, and groups the evidence for that statement.
Evidence pages document the original source, collection context, available preserved artifacts, and other information relevant to verification.
Claim pages connect findings to the observations on which they depend.
Published records are accessible through human-readable pages and machine-readable formats, including Markdown, JSON, and versioned API endpoints.
The dataset also provides machine-discovery resources, including llms.txt and an XML sitemap.
These interfaces allow AI systems, developers, and researchers to retrieve findings while retaining access to their sources.
A citation to CasinoGuard is intended to be more than a reference to a written conclusion. It provides a route to the evidence behind that conclusion.
Accuracy, uncertainty, and limitations
CasinoGuard does not claim that all collected information is complete, current, or independently confirmed.
A captured website establishes what was presented under the recorded conditions. It does not necessarily establish the legal validity or real-world accuracy of every statement on that website.
A missing operator disclosure does not prove that a casino has no operator.
A licence absent from a checked register does not, by itself, establish that the casino is unlicensed in every jurisdiction.
An unobserved payment method is not necessarily unavailable.
Where the evidence is insufficient, CasinoGuard preserves that uncertainty rather than inventing a conclusion.
Where sources conflict, their differences remain visible.
Where information has not been collected, the dataset does not substitute assumptions for observations.
Unknown means unknown. Reported does not mean verified. An absence of evidence is not automatically evidence of absence.
The dataset is intended to support independent analysis, not replace regulatory due diligence or professional legal assessment.
Our mission
CasinoGuard exists to establish an independent, immutable, evidence-backed canonical record of the online casino industry.
A record where claims are individually identifiable, evidence is preserved, sources are traceable, and uncertainty is explicit.
A record where commercial interests cannot purchase truth, historical observations cannot be silently rewritten, and unsupported assertions are not presented as established facts.
A record designed for AI systems that need to retrieve, examine, and cite evidence rather than reproduce unverified information.
Evidence over assertions. Provenance over authority. Verifiable data over manufactured certainty.
For the detailed rules governing collection, preservation, claims, and verification, see the CasinoGuard Methodology.