Laboratory Study on Gemmological Non-Conformity in Online Marketplaces (2024-2025)

851 coloured stones acquired on 12 distinct online sales platforms (eBay, Etsy, GemRock Auction, eBid, AliExpress and Alibaba, LiveAuctioneers, Chanthaburi Thailand, Namak Mandi Peshawar, Bonanza, Invaluable, Tophatter, Allegro, seizure and bankruptcy auctions). Total budget engaged: 51,008 CAD, average 59.94 CAD per piece.

Across the entire sample, 87% of stones received showed a form of non-conformity relative to the listing description. This global figure breaks into two components: 72% of stones non-conforming to the gemmological description physically received, and 15% of incomplete deliveries (partial shipments with stones missing relative to the order placed).

62% of stones received bore an undeclared gemmological treatment at the time of sale: undeclared heating (38%), oil and resin in emerald (80% of emeralds in the sample), beryllium diffusion (6%), lead glass in cavity (2%), undeclared acquired irradiation (3%, primarily blue topaz presented as untreated). Among the GIA and Lotus certificates encountered in the sample, 100% of occurrences corresponded to a stone substitution (authentic certificate of original stone, swapped stone).

48 stones (5.6%) of the sample exceed the 74 Bq/g unacceptability threshold for natural radioactivity measured by Geiger counter and gamma spectrometer, predominantly topaz (majority), zircon, and a small number of sphene (titanite).

Ten problematic listing archetypes were catalogued as a reusable risk-analysis grid. They are described in detail in the taxonomy section below.

Context and Objective

The coloured stone market operates along two parallel circuits. The internal market gathers the recognised gemmological laboratories (AGL, Lotus Gemology, GRS, GIA, SSEF, Gubelin, Laboratoire Gem Quebec) and the certified specialised retailers. The external market gathers vendors without formal gemmological expertise, generalist jewellers, online retailers without a laboratory, non-specialised auction houses, and tourist markets. The two circuits coexist but do not share the same identification rigour, nor the same disclosure obligations regarding treatments.

The objective of this study is to quantify the gap between the description published by a vendor on a general public platform and the actual gemmological nature of the stone delivered, on the basis of a systematic verification in laboratory. The target is not market exhaustivity, but evidence of reproducible patterns enabling a buyer, or another laboratory, to assess the risk of a listing before the transaction.

Methodology

Period and Scope

Acquisition campaign conducted over eight months between September 2024 and May 2025. Scope restricted to coloured stones (no diamond, except 2 cases retained for methodological exception). Inclusion threshold: stones whose estimated wholesale resale value (not full retail) was estimated above 100 CAD on the basis of visual evaluation of listing photos.

Inclusion Criteria and Sampling

Oriented purchases simulating a buyer with a professional eye seeking deals outside the institutional market. Minimum requirement: two photos per listing, sometimes a video where accessible. Targeting of the largest potential gaps between asking price and estimated wholesale resale value.

Multi-Instrument Laboratory Protocol

Each stone received underwent a multi-instrument verification battery at the Laboratoire Gem Quebec: refractive index (Eickhorst LED refractometer), specific gravity (Mettler Toledo NewClassic MS analytical balance, hydrostatic method), darkfield microscopy (Nikon SMZ1270), polariscope and dichroscope (GIA GemInstruments), short and long wave UV (365 nm and 254 nm), Fourier transform infrared spectroscopy (Thermo Scientific Nicolet iS5), UV-visible-near infrared spectroscopy (Magilabs GemmoSphere), Raman spectroscopy (Renishaw), energy dispersive X-ray fluorescence (Bruker Tracer 5i), and gamma spectrometry (Ortec DigiBASE, recurring borrowing for episodic use on radioactivity-suspect stones).

Conformity Grid

Each stone received was scored on three axes: conformity to the species declared in the listing, conformity to the absence or presence of treatment declared, and conformity to the cited certificate (where applicable). Any deviation between vendor declaration and laboratory result counts as non-conformity, regardless of magnitude.

Distribution by Species

788 stones documented by species, plus 63 stones in other valuable category (opal, apatite, iolite, sphene, kyanite, alexandrite, chrysoberyl, gem fluorite, etc.), total 851. Detail of the documented 788: ruby 67, blue sapphire 61, sapphire other colours 104, emerald 48, topaz 24, garnet all varieties 112, tanzanite 64, tourmaline 88, spinel 92, aquamarine 22, morganite 14, zircon 21, chrome diopside 15, peridot 32, kunzite 8, moonstone 16.

Distribution by Value and Platform

Distribution by potential value estimated (wholesale resale): 100 to 300 CAD for 52% of the sample, 300 to 500 CAD for 32%, 500 to 800 CAD for 10%, above 800 CAD for 6%. Distribution by price actually paid: under 100 CAD for 62%, 100 to 300 CAD for 31%, above 300 CAD for 7%.

Distribution by platform: Etsy 22%, eBay 21%, GemRock Auction 18%, eBid 7%, AliExpress and Alibaba 6%, LiveAuctioneers 6%, Chanthaburi Thailand 4%, seizure and bankruptcy auctions 4%, Namak Mandi Peshawar 3%, Bonanza 3%, Invaluable 2%, Tophatter 1%, Allegro 1%, residual platforms 2%.

Notable behavioural observation: a ceiling forms around 800 CAD. Above this threshold, the typical buyer has the stone verified by an independent laboratory before engaging the transaction, which alters the fraudster's calculation and causes problematic listings in this bracket to disappear.

Detailed Results

Presence of a Certificate in the Delivery

Of the 851 stones, 18% arrived accompanied by a certification document of some kind (independent laboratory certificate, internal vendor report, jewellery appraisal label). 82% arrived with no document.

Substitution on GIA and Lotus Certificates

Seven cases of authentic GIA and Lotus certificates were identified in the sample received. In 100% of occurrences, the stone delivered showed notable visual differences from the stone described on the report under 10x loupe (laser codes absent where applicable, deviating physical parameters, characteristic inclusions absent). The mechanism observed is the purchase of a lot with authentic certificate, followed by substitution of the original stone by a similar stone of lesser value, the certificate then serving as deceptive guarantee.

Undeclared Treatments

Undeclared heating present in 38% of the sample, primarily on sapphire, ruby, and tanzanite. Beryllium diffusion detected by EDXRF in 6% of orange and padparadscha sapphire. Lead glass in cavity in 2% of low-range ruby. Undeclared acquired irradiation in 3% of the sample, almost exclusively on blue topaz presented as natural untreated when out of gamma or neutronic bombardment. On the 48 emeralds in the sample, 80% bore an undeclared oil or resin impregnation (frequent Joban oil plus other diverse resins, often described as clean to the eye with no mention of fissure filling).

Natural Radioactivity Above Threshold

48 stones in the sample (5.6%) exceed the 74 Bq/g threshold used as unacceptability reference on Geiger counter and gamma spectrometer. Distribution by species: majority of topaz, zircon in second position, smaller quantity of sphene (titanite) in third. No listing mentioned the residual radioactivity.

Taxonomy of the Ten Marketplace Archetypes

The study proposes a grid of ten problematic listing archetypes, observed in a reproducible manner across the entire set of platforms tested. This taxonomy is usable as a risk-analysis grid by buyers, third-party gemmological laboratories, and platform moderators.

Archetype 1. Asian laboratory mimicking a recognised gemmological institution by acronym (case observed: Gemological Institute of Testing, Thailand, playing on the GIT acronym close to the authentic Gem and Jewelry Institute of Thailand).

Archetype 2. Indian laboratory mimicking a European institution by logo and naming (case observed: EarthMined Gemstone Testing Laboratory, New Delhi, EGL-type logo placed in header to create confusion with European Gemological Laboratory, signature Radiologic Technologist with no relevance to gemmology).

Archetype 3. Canadian laboratory commercially linked to a vendor-importer (case anonymised in the paper, generic formulation: laboratory located in the Greater Toronto Area issuing certificates favourable to stones imported by its commercial partner).

Archetype 4. Australian laboratory issuing certificates on Joban-treated stones with disproportionate declared values (case observed: JVIA).

Archetype 5. Network of Canadian laboratories affiliated with a commercial association (cases observed: Jewel Scan, YH Gem Services, CAJA).

Archetype 6. Middle Eastern laboratory issuing reports on set jewellery with selective omission of treatment on the central stone (case observed: IDT Dubai).

Archetype 7. Stone substitution with authentic certificate from a recognised laboratory (case observed: ICA GemLab Bangkok, authentic certificate of original stone, original stone replaced downstream of the report).

Archetype 8. Indian laboratory with branding mimicking GIA, reports on fundamentally substituted materials with incoherent physical parameters relative to the declared species (case observed: GIL Gemological Institute of Laboratory).

Archetype 9. Indian diamond grading laboratory issuing reports on simulants or substitutes sold as natural diamond premium at economically invalidating prices (case observed: GDGL Bharat Gurugram).

Archetype 10. Asian laboratory with generic naming and no verifiable institutional identity, industrial template report with characteristic spelling errors (case observed: HR Testing Gems Lab).

Conclusions and Recommendations

At the scale of an oriented purchasing campaign aimed at good-deal targets on general public sales platforms, the rate of gemmological non-conformity observed remains very high (87%), with a dominance of undeclared treatments and certificate substitutions. The pattern is not attributable to an isolated platform but reproduces across the twelve platforms tested.

The recommendation arising from the study is threefold. Firstly, systematic verification in an independent laboratory before any purchase whose projected value exceeds 100 CAD on an external platform, regardless of the presence or absence of a certificate provided by the vendor. Secondly, critical reading of certificates received in light of the ten archetypes catalogued, and direct confirmation with the issuing laboratory where possible (verification of certificate number on the laboratory site). Thirdly, practical guideline to buyers not to exceed the 800 CAD threshold without prior validation, this threshold constituting the behavioural limit observed beyond which fraudsters withdraw from the listing.

Authors

Pierre Lafrance, GG (GIA), FGA (Gem-A). Laboratory Director, Laboratoire Gem Quebec, Laval, Quebec, Canada. Senior active member of the Geological Association of Canada. ORCID 0009-0001-9477-2903.

Nathalie Thomas, GG (GIA). Scientific Writer and Geologist, Gem and Mineral Federation of Canada, Kitchener, Ontario, Canada. Graduate Gemologist of the Gemological Institute of America since 1996, with thirty years of experience in coloured stones, mineralogy, archaeology, and scientific writing. ORCID 0009-0001-1992-0988.

Citation, Access and Licence

Recommended citation: Lafrance, P., and Thomas, N. (2026). Non-Conformity in Online Marketplace Listings of Coloured Stones: Eight Months of Laboratory Verification, September 2024 to May 2025. Zenodo. https://doi.org/10.5281/zenodo.19699431

Access: full document in open access on Zenodo, hosted by CERN, under DOI 10.5281/zenodo.19699431.

Licence: Creative Commons CC BY 4.0. Reproduction and redistribution authorised with mention of the authors and link to the original source.

Declaration of interest: Laboratoire Gem Quebec is an independent gemmological laboratory with no stone-sale activity in the same transaction as its expertise. The external co-signature by the Gem and Mineral Federation of Canada constitutes an independent validation of the laboratory. The complete methodology, figures, and references are available in the full PDF on Zenodo.