The Provenance Handbook

There is a moment that happens at the bench perhaps once a week. The stone is out of its parcel paper and sitting under the daylight lamp, and whoever is sitting opposite me has stopped talking about colour. They have gone quiet, the way people do when a thought arrives that they were not expecting to have. And then they ask it: where did this actually come from?

It is the best question anyone asks me, and it is the one I find hardest to answer briefly. Because it is not one question. It is two questions wearing the same coat, and they have entirely different kinds of answer.

Two questions wearing one coat

The first question is geological. In what ground did this crystal grow? That is a question about aluminium and trace iron and metamorphic history, and a well-equipped laboratory can offer a considered professional answer to it, with instruments most of us will never operate. The answer will be a country, sometimes a region, and — as we will see in Chapter III — it will be an opinion rather than a fact, however confidently it is printed.

The second question is human. Whose hands dug this out, who bought it from them, at what price, and was anybody hurt or cheated along the way? No spectrometer answers that. Only a paper trail does, and paper trails in this trade are thinner than buyers imagine and thinner than sellers like to admit.

Almost every consumer guide I have read online squashes these two questions into a single adjective — ethical — and then treats a laboratory origin report as though it were the evidence. It is not. An origin report says where a stone probably formed. A provenance record says who touched it. A stone can have an immaculate origin report and no provenance record at all. It can also, less often, have a documented journey and an inconclusive origin. Neither one substitutes for the other, and confusing them is the single most common error in the way coloured stones are sold.

What this handbook will and will not do

What it will do: describe how a coloured stone genuinely moves from ground to shop window, with the numbers that are actually published rather than the ones that circulate in marketing copy. Explain what a laboratory can and cannot determine about origin, in the laboratories' own words. Take the traceability technologies seriously — DNA tagging, blockchain ledgers, registered rough — and then tell you precisely where each of them stops working, particularly for sapphire. Walk through the two mining regimes I know best, Sri Lanka and Australia, including the parts that do not flatter either. Set out the rules that now bind a dealer in this country, some of which changed in the past two months. And finish with the questions I would ask a seller if I were the one buying.

What it will not do: tell you that a stone is a good investment, or that any category of gem will hold or grow its value. Nothing in this handbook is investment or financial advice, and where I discuss market values or premiums it is to help you judge a price today, not to forecast one. It also will not tell you that there is a clean shelf and a dirty shelf and that we happen to stand on the clean one. That framing is false, and I think most buyers can smell it.

The house position, stated up front

So that you can read the rest of this knowing where I stand: my family has worked in the Ceylon gem fields across three generations since the 1960s, and I assess every stone that passes through this house myself, here in Sydney. We sell only natural, earth-mined stones. That is a preference, not a verdict — laboratory-grown gems are a legitimate product, honestly made, and I have no interest in sneering at them; they simply are not what we do, and we say which is which on every listing.

On treatments, my position is unfashionably boring. Heating a sapphire is a legitimate, ancient and disclosed practice, and a heated stone honestly described and honestly priced is a perfectly good thing to own. The sin has never been the treatment. The sin is silence.

And on provenance, the same principle applies with more force, because provenance is where silence is easiest to get away with. Where we know a stone's journey, we write it down. Where we do not, we say so — in writing, on the record, rather than reaching for a comforting adjective and hoping nobody asks a second question.

— Anna, FGAA
Ministry of Gems, Sydney

If you have a mental image of gem mining, it is probably wrong, and the way it is wrong matters for everything that follows. Most people picture a company: a fenced site, a haul truck, a payroll, an annual report. That picture is accurate for a small minority of coloured-stone production and for almost none of the sapphire trade.

Eighty per cent of it comes out of a hand-dug hole

Artisanal and small-scale mining — ASM, in the acronym-heavy language of the development sector — accounts for roughly 80% of global coloured gemstone production, according to the World Bank and Pact's Delve data platform, current to 2024. The same sector supplies a far smaller share of other minerals: 10–25% of gold, 15–20% of diamonds, 20–25% of tantalum and tin. Coloured stones are the outlier. They are the mineral category most thoroughly in the hands of individuals.

For sapphire specifically, the World Bank's estimate is that 80% of global production comes from artisanal groups, a figure quoted in the Coloured Gemstones Working Group's 2021 white paper Glimmers in the Shadows. GIA's own literature review puts the coloured-stone figure at 75–80%. And from the other direction, the same white paper estimates that large-scale mining accounts for only about 20–30% of non-jade coloured gemstone production, concentrated in high-value ruby and emerald, with limited room to grow because very few coloured-stone deposits are large enough to justify industrial capital.

The hands in between

A stone typically leaves the miner via a local trader who travels out to the mine site. From there it moves to a town buyer or market — Beruwala, in the Sri Lankan case — then to a national hub such as Colombo, "changing hands several more times on the way to reach an international trader or buyer", in the white paper's phrasing. Then it goes to a cutting centre, quite possibly in a different country from the one it was mined in, then very often to a treatment specialist, then to wholesalers, manufacturers and finally retailers.

You will see the claim that a coloured stone changes hands "ten to fifteen times" repeated across the internet. I am not going to repeat it, because I cannot find a primary study behind it; the best-sourced framing available is the Coloured Gemstones Working Group's own, which says "several times" and "as many as ten times" among local traders before the stone progresses further down the chain. The round number is retail copy. The vaguer number is research. I would rather give you the research.

Two features of this chain do most of the damage. The first is informality: trading near mine sites happens largely without paperwork, and largely in cash. The second is an information gap. Even professional gemmologists' valuations of the same piece of rough can vary by up to 30%, which tells you how structurally disadvantaged a miner is at the first sale, when the only person in the transaction who knows the international market is the one buying.

Where traceability structurally dies

There are two places in this system where mine-to-market traceability does not merely fail; it cannot exist.

The first is melee. Stones under 0.20 carat are sorted and sold in parcels by size and quality, and those parcels aggregate stones from all over the world. A parcel of 1.5 mm blue sapphire is not a batch from one mine that has lost its paperwork. It is, by construction, a mixture. Any seller offering you a fully traced pavé of small accent stones is describing something the trade does not currently produce at scale.

The second is time. Gemstones are effectively imperishable, and it is rare for a gemstone to have only one final owner; stones re-enter the market through estates, auctions, re-cutting and re-treatment "long after the mine that they came from has closed". Put a figure on that and it becomes startling. The gemmologists Laurent Cartier and Vincent Pardieu estimated in a 2012 piece for the National Geographic Society that fewer than 2% of gemstones in circulation were mined in the last two years.

Sit with that. If the estimate is anywhere near right, then "mine-to-market traceability" is structurally unavailable for the overwhelming majority of coloured stones on the market at any given moment — including essentially all estate, antique and recut material. I have never seen a consumer sourcing guide say this. It is the most important single fact in the whole subject, and it is missing from all of them.

Smuggling, cutting, and the word "Ceylon"

Small, valuable, easily concealed goods move across borders informally. GIA's review notes an estimated US$60 million worth of emeralds smuggled out of Zambia every year. In Madagascar, a US dealer quoted by The Guardian in April 2017 estimated that about 70% of the country's sapphire market was controlled by Sri Lankan buyers who moved the gems back to Sri Lanka to be cut and exported, with perhaps US$150 million of sapphire leaving Madagascar annually — a figure the same article concedes is unknowable, precisely because the industry is not well regulated.

You can see the same pattern in trade statistics. In 2016, Myanmar recorded just over US$1.8 million of uncut precious and semi-precious stone exports to Thailand. Thailand recorded receiving US$4.7 million.

Here is why this belongs in a buyer's handbook rather than a policy paper: smuggled stones are routinely re-badged with the origin of the country they are smuggled into. African sapphires are frequently moved into Sri Lanka and sold as "Ceylon" sapphires, because African material is much cheaper. The word "Ceylon" on a listing is therefore a claim, not a certificate — and a claim with a well-documented incentive behind it. Ask what it rests on.

One more link deserves naming, because it is the one nobody photographs. The Solidarity Center estimates that 80% of all coloured stones on the market are cut in cutters' homes or small workshops, and an estimated 30% of gemstone grinders will die of silicosis. Those figures, cited in GIA's 2016 review, describe the stage of the chain furthest from the buyer's imagination and closest to the finished stone in their hand.

This chapter is easily misread as an attack on gem laboratories. It is the opposite. They have been telling the trade the truth about origin determination for two decades, in print, on their own reports. The trade has simply chosen not to pass it on.

The sentence every laboratory prints and nobody reads

GIA put it plainly in the Winter 2019 issue of Gems & Gemology, in an article titled "The Geographic Origin Dilemma": "all gem labs emphasize that geographic origin determinations on reports are opinions — they are not indisputable facts. Because they are opinions, there will sometimes be differences in reports from different organizations."

The Swiss Gemmological Institute has said the same for longer. Dr Michael Krzemnicki of SSEF wrote in InColor back in Winter 2007 that "an origin determination in a printed gemstone report is always an opinion, based on the existing knowledge of the gemmologist(s) in the laboratory that prepared it. Different laboratories may have different philosophies of origin determination, resulting in different origin determinations for the same gemstone."

GIA notes that every major laboratory prints a comment to this effect on its reports. Go and look at the small type on any origin report you own. It is there.

And "inconclusive" is a real outcome, not a failure. GIA again: "No matter how hard we try, sometimes we simply cannot tell where a stone came from … In such cases, we will say the origin is 'inconclusive,' which is never a popular conclusion. Regardless, in such situations it is the truth." I have a great deal of time for a laboratory willing to write that sentence about its own paid work.

Why the premise itself is shaky

Origin determination rests on an assumption: that stones from different deposits have measurably different properties. GIA states flatly that this "is not always true". Deposits in different countries can produce gems with near-identical signatures — "some blue sapphires from Myanmar are virtually identical to some blue sapphires from Sri Lanka" — while deposits inside a single country can differ radically, as with Madagascar, where most blue sapphire is metamorphic but the Ambondromifehy material is basalt-hosted.

Criteria also expire, and this one is instructive. Orange fluorescence in blue sapphire was for years taken to indicate a Sri Lankan origin. Further research established that it occurs across several marble-type, low-iron metamorphic deposits, and GIA now describes it as "mostly unhelpful in separating origin". A diagnostic that was good practice in one decade became a mistake in the next. Any origin opinion is therefore also dated — it reflects what was known when it was written.

Even the coarse split between metamorphic and basalt-related sapphire, which is the first fork in the road, is imperfect: GIA's own methodology paper concedes "there is overlap and trace element chemistry alone cannot fully separate these two groups". And metamorphic blue sapphire is the hardest case of all — "even without the arrival of Madagascar sapphires, it is not always possible to separate the three classical sources with 100% confidence". New deposits keep breaking old rules, and SSEF names Mozambican and Nigerian copper-bearing tourmalines and the Himalayan dolomite-marble rubies of Afghanistan and its neighbours as material that is not easily distinguished from established sources.

Twenty-two thousand stones, and still counting

Since 2008, through field expeditions on six continents, GIA has accumulated more than 22,000 coloured-stone reference samples with known origins. The published blue-sapphire methodology rests on trace-element data from 606 metamorphic sapphires — 124 from Sri Lanka, 263 from Madagascar, 219 from Myanmar — and 342 basalt-related sapphires from Nigeria, Australia, Thailand, Cambodia and Ethiopia. The analytical toolkit runs from microscopy and UV-Vis-NIR spectroscopy through ED-XRF and LIBS to laser ablation ICP-MS, with trace elements now measured down to parts per billion in some cases, all of it requiring continual recalibration against known standards and operators with advanced degrees.

One practical note for buyers: inclusions are supporting evidence, not proof. GIA advises that inclusion features "should be used as supporting evidence in addition to chemical analysis as these features often overlap significantly". If a seller's entire origin argument is a photograph of a silk cloud, that is an argument, not a determination.

The value gap, and what it tempts people to do

None of this would matter much if origin were merely interesting. It is not merely interesting; it is money. GIA observes that a lower-quality Kashmir sapphire may be more expensive than a much finer Sri Lankan sapphire on provenance alone, and that a fine classical Kashmir stone can sell for many times more than a Madagascan sapphire of exceptional quality and size.

Where a word on a page multiplies a price, there is pressure on the word. GIA says so outright: "The fact that certain origins can add value creates a situation where some dealers will sell a stone with a laboratory report that they know is incorrect." One dealer told the authors, in a line I have never quite been able to forget, "It does not matter where a stone actually come from. It only matters where the labs say it comes from." The article also records the trade term "lucky Burma" — Sri Lankan sapphires close enough to Burmese material to be knowingly sold as Burmese.

The disagreements are documented, too. GIA cites a 5.22 ct ruby carrying four different identification reports, with country of origin given as either Afghanistan or Vietnam. When reports conflict, auction houses now often simply list all of them in the catalogue and let the buyer weigh the evidence — which is, I think, the correct professional response, and a quietly radical admission about the state of the science.

So: buy origin reports, read them, keep them. Just hold them for what they are. A good report from a good laboratory is the most rigorous opinion available on a genuinely hard question. It is not a birth certificate, and it never claimed to be.

Origin determination reads the stone. Traceability documents the journey. They are solutions to different problems, and the gap between them is precisely where the last decade of technology has been aimed.

The gap is easy to describe. A laboratory can identify geological properties typical of a region, but that cannot be resolved down to a specific mine within a given geological region. So if you want to know not just "Sri Lanka" but "this pit, this licence, this co-operative", the stone itself will not tell you. Something has to be attached to it, or written down about it.

Tagging the stone itself

The most elegant answer belongs to the Gübelin Gem Lab, whose Emerald Paternity Test launched in 2017. Rough emerald is immersed in a bath of DNA-based nanoparticles at the mine site; the particles lodge in the stone's fissures deeply enough to survive cutting and clarity enhancement, and at around 0.0001 mm across they are invisible to the naked eye and under an optical microscope, so grading is untouched. Haelixa, an ETH Zurich spin-off founded in 2016, works on the same principle with synthetic DNA markers read back by PCR at its laboratory in Kemptthal, Switzerland, and has partnered with Gübelin's Provenance Proof initiative on emerald tracing since 2017.

It is genuinely clever engineering. And here is the limit that a sapphire house has an obligation to state loudly: it does not transfer to ruby and sapphire. Most corundum is heat treated, and heat treatment destroys the DNA nanoparticles. The emerald model therefore cannot simply be extended across the corundum trade.

The ledger problem: immutable is not the same as true

The Provenance Proof Blockchain was launched by Gübelin in partnership with Everledger, deliberately opened to all industry stakeholders, with both data contribution and access free of charge. That openness was the right instinct — a proprietary provenance ledger owned by one house is a marketing asset, not an industry solution.

But two cautions apply, and the first is structural. A blockchain guarantees that a record cannot be altered once entered. It guarantees nothing whatsoever about whether the original entry was accurate. If a stone is swapped, mislabelled or fraudulently registered at the point of entry, the ledger will faithfully and permanently record the fraud, with a lovely cryptographic seal on it. Closing that gap needs robust physical-to-digital binding, independent auditing and oversight at the moment of entry — all of which live outside the chain. The academic literature on blockchain in supply chains is consistent on this point, and there is a further critique worth knowing: research on "digital extraction" argues that these systems can reproduce rather than dissolve existing power asymmetries, and can exclude the artisanal producers they are meant to help.

The second caution is commercial. Everledger — which had provided tracking for GIA, Brilliant Earth and Provenance Proof, and had raised over US$51.7 million — entered voluntary administration after a second funding tranche failed to materialise; staff were given notice on 31 March 2023 and administrators were appointed on 24 April 2023. Provenance data that lives entirely inside one venture-funded platform is only as durable as that platform's balance sheet. If you are ever told a stone's provenance is "on the blockchain", the useful follow-up question is: whose, and what happens to my record if they close?

What diamonds got that coloured stones did not

It is instructive to look sideways at the diamond trade, where traceability has consolidated in a way coloured stones have not. De Beers' Tracr platform now has more than five million rough diamonds registered at source, representing roughly two-thirds of De Beers' rough production by value, and single country of origin has been available on the platform since January 2025. On 29 May 2026, GIA acquired a 30% shareholding in Tracr, which De Beers framed as the platform's evolution into an independent, industry-wide platform.

That is real institutional weight. But note the shape of it: one dominant producer, a small number of mines, a highly concentrated rough market, and a single registration point. Coloured stones have the opposite structure — roughly 80% artisanal production across scattered diggings that, in GIA's words, appear and disappear virtually overnight, and for most of which there are no recorded production estimates at all. The reason coloured stones have no Tracr is not that nobody has tried. It is that the supply chain has no chokepoint at which to install one.

So what can actually be documented today

Documented, verifiable, working models exist. Moyo Gems, launched in 2019, works with women artisanal miners in Tanzania's Umba Valley and Kenya's Taita Taveta County, with miners receiving 95% of the export value of their gemstones and local brokers 5% for sorting, valuation support and facilitation — a chain that is short enough, and deliberate enough, to be documented end to end. It proves the model is possible at a certain scale. It does not scale to the whole market, and nobody involved pretends it does.

Below that, a great deal is achievable with ordinary records rather than exotic technology: purchase documentation naming the actual seller and date, laboratory reports issued before the stone changed hands again, treatment disclosure in writing, photographs and weights recorded at each stage, and a dealer willing to put their knowledge and their ignorance in the same document. That is unglamorous. It is also, for most stones, the only provenance that genuinely exists — and it is the standard I would hold any seller to, including this one.

This is the ground my family has worked across three generations, so treat what follows as informed rather than neutral, and check me against the sources I name. I have tried to be harder on Sri Lanka here than a marketing page would be, because the defensible parts of the story only stay defensible if the rest is named too.

What the law actually forbids

Gem mining in Sri Lanka is licensed by a single authority, the National Gem and Jewellery Authority, operating under the National Gem and Jewellery Authority Act No. 50 of 1993. There is no patchwork of state regimes to navigate; there is one body, and it issues the licences.

The striking feature — the one I would put first if I had to explain Ceylon sourcing in a sentence — is that the Sri Lankan government has banned the use of heavy mechanised mining methods in gem mining. A British Geological Survey field study conducted for the UK's Department for International Development recorded it directly: large mechanised equipment such as bulldozers and excavators is not used for mining and is banned by the government. The stated rationale is twofold, and both halves are worth hearing. Mechanisation would strip an irreplaceable resource quickly; and it would destroy the rural livelihoods built around working it slowly.

The same study records that the processing side is equally low-technology: no mechanised jigs or sieves are used to process the gem-bearing gravel, the illam, and no hydraulic hoses or pressure jets are used. The washing and picking is done by hand and by eye. If you have seen footage of high-pressure water mining tearing apart a riverbank elsewhere in the world, that is precisely the practice this regime forecloses.

Mechanisation is not absolutely prohibited, and I would rather say so than overstate the case. The NGJA permits it in narrow circumstances — particularly rich deposits that would otherwise attract illegal operators, or ground where gem concentration is too low for viable pit mining. But a mechanical gem mining permit runs for one year only, is available only on land already holding a general gem mining licence of one acre or more, and issues only with the recommendation and approval of the NGJA's Excavation Engineer and Functional Officer.

Two further structural details matter. No gem mining is permitted in State Parks and wildlife reserves. And land-type consent is layered on top of the licence itself: paddy land requires a recommendation from the Department of Agrarian Development, temple land from the Department of Buddhist Affairs, government grant land from the Divisional Secretary, Land Reforms Commission land from the Commission, and Mahaweli Authority land from the Authority. A gem licence does not override the question of whose ground it is.

The deposit that does the work

The mechanism I find most quietly effective is financial. Alongside the fees — an application fee of Rs. 800 plus tax, an initial fee of Rs. 1,500 plus tax for single ownership or Rs. 5,000 plus tax for a partnership, and an annual mine fee of Rs. 8,000 plus tax — the licensee lodges a refundable guarantee deposit of Rs. 10,000. Licences run for one year.

That deposit is the restoration mechanism. Two months before the licence expires, the licensee applies to have it released. The NGJA inspects the land, and the money comes back only once proper pit closure and rehabilitation are confirmed. Shafts and open pits must be filled after mining, the stated reasons being groundwater contamination, landscape damage and mosquito breeding. Sri Lanka's Export Development Board claims the country therefore has comparatively few abandoned unfilled pits relative to other gem-mining countries; that is an industry body's assessment rather than an audited statistic, and the reporting below puts pressure on it.

The licence process is also publicly contestable, which surprises people. After a Regional Development Officer inspects the land, a notice is displayed on the site for fourteen days inviting objections, and the licence issues only if no objection is upheld. Your neighbours get a vote, in effect, before anyone digs.

Who is in the pit, and on what terms

Labour is organised through informal co-operatives with a defined skill hierarchy — labourers, washers and sorters, and at least one experienced gem picker whose eye is the most valuable instrument on site. Returns are split by formula rather than paid as a wage: a tenth of income to the water pump, a fifth of the balance as the land share, a tenth of the balance to the licence holder, before costs. Pit miners are effectively shareholders. The NGJA also states that licensed gem mining carries mandated life insurance, accident and disability compensation, and educational scholarships for miners' children.

Where the regime does not hold

Now the other half, because a sourcing story that only contains the good parts is advertising.

Enforcement is imperfect, and Sri Lankan reporting says so. Journalism published through the Earth Journalism Network documents abandoned pits in the gem districts — usually in paddy fields — left open for three years and in some cases a decade, into which livestock and people have fallen. The NGJA's position is that pit closure is the licensee's responsibility. A refundable deposit is an elegant incentive; it is not an inspector standing in a field.

The environmental cost is real even under hand methods. Peer-reviewed work on small-scale gem mining in Sri Lanka associates tunnel, backhoe and river mining with water contamination, soil erosion, deforestation, biodiversity and habitat loss, and nutrient depletion. "Less damaging than the alternative" is not the same as "harmless", and I try not to let the first phrase quietly become the second.

There is also an illegal sector. The 1999 BGS/DFID study cited estimates of up to 15,000 illegal gem-mining operations and concluded there were at least as many illegal as legal operations. I quote that figure with an explicit caution: it is from 1999, and I have found no current equivalent, so it should not be presented as today's position. It does establish that the problem existed at serious scale within living memory, and no one has published a number showing it has gone away.

Finally, the national statistics do not agree with each other. Reported sector export values include US$500m and US$477.7m for 2023, US$381.9m for 2024, and US$334.13m on the Export Development Board's own page, while Sri Lankan press has reported unaccounted-for gem exports exceeding US$1 billion — larger than the entire declared trade. I am not going to arbitrate between those figures. I will say that when the informal estimate exceeds the formal one, the honest description of the sector is "partly documented", and any seller who tells you otherwise is guessing with confidence.

My summary, then, is a two-part sentence I try never to shorten: the Sri Lankan regime is comparatively benign by design, and compliance with it is uneven. A gem house is entitled to claim the first. It is not entitled to claim the second on a stone's behalf without evidence about that particular stone.

Australian material has a marketing advantage it did not ask for: buyers assume that because it comes from a wealthy country with a functioning regulator, it must be fine. The assumption is broadly reasonable and, as the regulator's own reports demonstrate, not automatically true. Both halves of that sentence belong in a handbook.

Sapphire: New England and the Central Queensland Gemfields

Australian sapphire occurs chiefly in alluvial and placer deposits, concentrated by weathering and fluvial transport rather than mined from primary rock. Geoscience Australia identifies two principal regions: the New England region of New South Wales, around Inverell and Glen Innes, and the Central Queensland Gemfields at Anakie, Rubyvale, Sapphire, The Willows and Glenalva. Commercial corundum mining still occurs at Anakie and Lava Plains in Queensland and in New England. Geoscience Australia notes that the Kings Plains area near Inverell has some of the richest deposits of gem-quality sapphires ever mined — a sentence that tends to surprise people who think of sapphire as an exclusively Asian story.

It once was a very large story indeed. In the 1980s Australia supplied approximately 70% of the world's sapphire, mostly from New South Wales. That share later fell to about 20–30% as mechanised mining expanded internationally. Australian sapphire did not become worse; the rest of the world simply industrialised faster.

Method matters here, and it is a fair, sourced point of contrast rather than a boast. Australian commercial operations use open-pit mining of gravels with jackhammers, excavators or shovels. Geoscience Australia observes that some countries use high-pressure water mining, and states that "gem buyers boycott companies using this practice" because of the environmental damage it causes. That is a government agency, not a marketing department, describing a real distinction in extraction method.

Opal: 95% of the world, dug by individuals

Australia produces 95% of the world's precious opal, has done so as the major producer since 1932, and opal is the country's official national gemstone. The principal fields are Lightning Ridge and White Cliffs in New South Wales; Coober Pedy, Andamooka and Mintabie in South Australia; and a boulder-opal belt running more than 700 km through Queensland, taking in Quilpie, Yowah and Winton.

Here is the detail people find hardest to believe. Geoscience Australia characterises Australian opal mining as remaining "in the hands of individuals who work within small mining leases or claims". Australia's opal sector is itself artisanal — not in the developing-economy sense, but structurally: small operators, small claims, small capital. Methods range from historical shaft-sinking and tunnelling with hand tools and winches, through modern nine-inch auger drilling and one-metre-diameter vertical shafts, to Queensland boulder-opal operations where bulldozers and 20- to 40-tonne excavators are widely used. The last of those is genuinely industrial, and worth knowing if the environmental footprint of your opal matters to you.

What the rules actually require

The New South Wales framework is unusually specific, and the specificity is the point. An opal mineral claim must be square and no larger than 50 m × 50 m — 2,500 square metres — with larger claims up to two hectares available only for reward claims or puddling operations. A person may hold no more than two mineral claims at any one time. You cannot quietly accumulate a field.

A rehabilitation security deposit is required, usually set at A$700, and is returned only after the small-scale title has been rehabilitated to the required standard. The standard at Lightning Ridge is prescriptive to a degree I rather admire: disturbed land must be left "safe, stable and self-sustaining"; shafts backfilled with mullock heaped over the shaft centre no more than one to two metres high; auger holes fully backfilled with a central mound of about 300 mm within a circle no more than a metre across, topped with 50 mm of red gravel; trenches fully backfilled with topsoil respread and cleared vegetation returned to assist revegetation. Topsoil must be stockpiled separately from mullock before mining begins. That is not aspirational language. That is a specification.

What the regulator actually found

And now the part that Australian sellers, including this one, are obliged to sit with.

Lightning Ridge hosts over 3,000 small mining claims. In July 2017 the NSW Resources Regulator ran a proactive compliance operation there: 19 inspectors examined 124 claims. Of 61 expired or cancelled claims, only 14 were compliant with rehabilitation requirements. Of 63 active claims, only 12 were compliant. Sixty-eight shaft or auger holes were not secured to guideline. The regulator's own summary sentence is the one to remember: "Overall compliance levels were found to be poor."

The lesson generalises well beyond opal. A jurisdiction claim is not a mine claim. "Australian" tells you which rules applied and which regulator had authority. It does not tell you whether the particular hole your stone came out of was one of the compliant ones. Australian origin narrows the risk set considerably. It does not close it, and any seller who implies otherwise is selling you a flag rather than a fact.

Adjectives are cheap and rules are not, which is why this is the chapter I would read first if I were assessing a seller rather than a stone. What a dealer is legally required to do tells you more about them than what they choose to say.

OECD: due diligence is a process, not a badge

The reference framework for responsible mineral sourcing is the OECD's Due Diligence Guidance for Responsible Supply Chains of Minerals from Conflict-Affected and High-Risk Areas, first published in 2011, with the third edition dated 2016. The OECD is candid about its status: "While not legally binding, the Recommendation reflects the common position and political commitment of OECD members and non-member adherents."

Two things about it matter to a coloured-stone buyer. First, scope. The third edition explicitly extended the Guidance to all minerals, removing language that had been read as limiting it to tin, tantalum, tungsten and gold. Coloured gemstones are therefore squarely in scope, and any dealer who tells you the OECD framework is a conflict-minerals thing that does not apply to sapphire is working from a decade-old understanding.

Second, shape. The Guidance sets out five steps: establish strong company management systems; identify and assess risk in the supply chain; design and implement a response strategy; carry out independent third-party audit of supply chain due diligence; and report annually. Notice what that structure implies. Due diligence is an ongoing, proactive and reactive risk-management process — not a certificate a product either has or lacks.

The Kimberley Process does not cover your sapphire

The Kimberley Process Certification Scheme was launched in November 2002 at Interlaken by 37 countries and entered into force in 2003. It defines conflict diamonds as rough diamonds used by rebel movements or their allies to finance conflict aimed at undermining legitimate governments. It has 60 participants representing 86 countries, and states that conflict diamonds now comprise less than 0.1% of worldwide rough diamond production.

And here is the scope. The KP regulates trade in rough diamonds only — HS codes 7102.10, 7102.21 and 7102.31. Polished diamonds fall outside its requirements. Coloured stones are not in it at all. There is no Kimberley certificate for a sapphire, there never has been, and if one is ever waved at you, something has gone badly wrong.

Its limits are contested even within diamonds. GIA summarises the scheme as extending "only to rough diamonds being used to fund rebel militia efforts to overthrow a legitimate government, not to other human rights issues" — which excludes state actors and excludes labour abuse by definition. Global Witness, a founding civil-society participant, withdrew in 2011, citing failures over Côte d'Ivoire, Venezuela and Zimbabwe.

Greenwashing is now an enforcement matter

In Australia, environmental and sustainability claims are policed under the Australian Consumer Law, and on 12 December 2023 the ACCC published final guidance, "Making environmental claims", built on eight principles: make accurate and truthful claims; have evidence to back them up; don't hide or omit important information; explain conditions or qualifications; avoid broad and unqualified claims; use clear language; ensure visual elements do not mislead; and be direct about your sustainability transition.

The scale of the problem is documented. In an internet sweep launched in October 2022 and reported in March 2023, the ACCC reviewed 247 business websites and found 57% had made concerning claims about their environmental credentials. More than half. When you next read a jewellery sourcing page, that is the base rate you are reading against.

The equivalent regime in the United States is the FTC's Jewelry Guides, revised in July 2018, whose operative principle is that origin claims which cannot be substantiated are not permitted. Principle five of the ACCC list — avoid broad and unqualified claims — is the one most jewellery copy fails, and "ethically sourced", standing alone with nothing behind it, is the textbook example.

AUSTRAC, and what changed on 1 July 2026

Parliament passed the AML/CTF Amendment Bill on 29 November 2024, amending the Anti-Money Laundering and Counter-Terrorism Financing Act 2006. Under those Tranche 2 reforms, dealers in precious stones, metals and products have been regulated by AUSTRAC from 1 July 2026. AUSTRAC opened enrolment on 31 March 2026, noting the number of businesses it regulates would grow from around 19,000 to close to 100,000 nationwide — changes its chief executive, Brendan Thomas, called "the most significant overhaul of Australia's AML/CTF framework in more than 20 years".

The designated service is precise. It covers buying or selling precious metal, precious stones or precious products in the course of carrying on a business, where the purchase involves the transfer of physical currency or virtual assets with a total value of not less than $10,000 — whether in a single transaction or in several that are "linked or appear to be linked". Two limbs get missed. Card and bank transfer sales fall outside the designated service entirely. And sub-threshold cash sellers still carry an obligation, because businesses must have processes to monitor for linked transactions that could reach the threshold.

Core obligations from 1 July 2026 are to enrol with AUSTRAC, implement an AML/CTF program, appoint a compliance officer, conduct initial customer due diligence, report suspicious matters and keep records, with enrolment required within 28 days of first providing a designated service. The Act provides civil penalties of up to 100,000 penalty units per contravention for a body corporate, and the Commonwealth penalty unit rose to $364 for offences committed on or after 1 July 2026 — I give you the penalty-unit figure rather than a dollar total, because the totals in circulation use the older value.

Why does this belong in a buyer's handbook? Because it is a free, checkable signal: a dealer who can talk fluently about their AUSTRAC enrolment, their customer due diligence and their record-keeping has had to build systems, and a dealer who looks blank in August 2026 has not. This is general information about the regime rather than legal or compliance advice.

One last thing before we leave the rules, because it bears on all of them. Certification is not a substitute for due diligence, and the best evidence comes from Human Rights Watch, which in November 2020 assessed 15 major jewellery and watch companies with over US$40 billion in combined annual revenue — around 15% of global jewellery sales. It found the Responsible Jewellery Council's "governance, standards, and system of audits are flawed, allowing companies to be RJC-certified even if they fail to meet basic human rights standards". Only 4 of the 15 reported tracing some mined gold to origin, and as of July 2020 only 106 of 1,256 RJC members — 8% — had adopted the voluntary Chain of Custody Standard, which does not require tracing to a mine of origin, only to countries.

HRW's structural criticism is that the RJC board is composed solely of industry representatives, that auditors are accredited by the RJC itself, and that audit reports are not made public. The RJC has publicly responded to and disputed that characterisation, and you are entitled to read both. HRW's own conclusion is blunt: "only mandatory human rights due diligence rules — laws — will create a level playing field."

I have been putting off this chapter through seven others, because it is the one where I have to be rude about my own industry, and by extension about the temptation this house shares with every other house in it.

The category error at the heart of the word

Start with the cleanest available statement of the problem, from GIA's 2016 ethics review: the industry's effort has "primarily focused on traceable chains of custody rather than sustainable actions", following Hilson's analysis, with the consequence that answers pertaining only to traceability may be "less than satisfying" to a customer asking about fair trade.

That sentence is the whole chapter in miniature. Traceability is not ethics. They are correlated, and people treat correlation as identity.

Consider the two failure modes. A stone can be fully traceable and unethical: you know precisely which mine it came from, and that mine underpays, or works children, or has never rehabilitated a pit in its life. Documentation records conditions; it does not improve them. And a stone can be ethical and untraceable, dug by a licensed co-operative that insures its people and then absorbed into a parcel three transactions later. The second stone is morally superior and commercially disadvantaged, and any market that rewards paperwork over practice will keep producing that outcome.

The OECD framing is the corrective. Responsible sourcing is a risk-management process — proactive, reactive, ongoing — not an adjective that attaches to an object. When you read "ethically sourced" on a product page, the honest translation is usually "we believe this, and here is the process by which we came to believe it", and the seller has simply skipped the second half.

The critique from inside the trade

GIA's 2016 ethics review records that "a number of skeptics consider this movement 'greenwashing,' an environmentalist-inspired marketing scheme to get consumers … to pay top dollar for an ultimately meaningless designation". It quotes Townsend's harder structural point, that corporate social responsibility "is, at best, only a partial solution — one which can be misused to create an illusion of responsibility". And it records Greg Valerio's description of a "cul-de-sac of inertia", in which the same industry figures talk to each other about responsibility rather than to consumers.

I do not think the sceptics are entirely right. I do think anyone selling gems should be able to recite their argument from memory, because the discipline of being able to state the case against yourself is most of what separates a claim from a slogan.

Two counter-arguments deserve equal airtime, and both come from the same GIA review. Boycotting is not the ethical answer: most experts agree that if consumers stop buying, artisanal miners bear the cost first and hardest. And deleting intermediaries wholesale is not the answer either, because middlemen provide market access, sorting and transport that miners cannot supply for themselves. "Buy nothing" and "buy only direct" both sound principled and both, at scale, hurt the people they claim to protect.

It is also worth keeping the environmental comparison honest in both directions. GIA notes that coloured gemstone mining is generally less environmentally hazardous than gold mining, because chemicals are not used and digging generally takes place within about 10 metres of the surface. In the same breath it stresses that considerable uncertainty about coloured-stone impacts remains, and that without land reclamation, soil degradation, deforestation and habitat harm are "inevitable". Shallow and chemical-free is a genuinely better starting point. It is not a clean bill of health.

What actually works, and at what scale

Moyo Gems, launched in 2019, works with women artisanal miners in Tanzania's Umba Valley and Kenya's Taita Taveta County. Miners receive 95% of the export value of their gemstones; local brokers receive 5% for sorting, valuation support and facilitation. Its partners include the Tanzanian Women Miners Association, AWEIK, Nineteen48, ANZA Gems and, formerly, Everledger. Set that 95% figure against the earlier finding that expert valuations of the same rough can differ by up to 30%, and you can see exactly which problem it is solving: not just the split, but the information asymmetry at the first sale.

The Coloured Gemstones Working Group, formed in 2015 by Tiffany & Co., Swarovski, Richemont, LVMH, Kering and Gemfields, launched a free Gemstones and Jewellery Community Platform in April 2021, built around ten sustainability commitments with free learning resources covering more than 40 sustainability topics. Free and open is the right architecture. Whether the commitments bite is a fair question, and one the sceptics above would press.

Lab-grown, recycled, and the arguments people reach for

Laboratory-grown stones are a legitimate product. They are real gem material, honestly manufactured, and when they are clearly disclosed as laboratory-grown there is nothing whatever wrong with buying one. I am not going to sneer at them, and I would ask you to be sceptical of anyone in my trade who does — the sneering usually tells you more about the seller's margins than about the stone. Our position is simply stated: we sell only natural, earth-mined stones. That is what this house is for. It is a choice about what we do, not a verdict on what anyone else does.

Recycled and estate stones are the other alternative, and they are far more common than people realise — recall that fewer than 2% of stones in circulation were mined in the last two years. An estate stone has no new mining footprint at all, which is a real and honest advantage. It also, almost always, has no provenance whatsoever. You are trading one kind of unknown for another. That is a perfectly reasonable trade to make with your eyes open, and an unreasonable thing to be sold as "traceable".

So here is how I would read a sourcing page, mine included. Look for numbers, dates and named sources. Look for the sentence where the seller states what they do not know. Look for whether "ethical" is doing work that "here is our process" should be doing. And notice that broad, unqualified claims are precisely what the ACCC's fifth principle warns against — so a page full of them is telling you something about the seller's care, quite apart from telling you nothing about the stone.

Everything above was context. This is the part you can use on a Saturday afternoon, with a seller in front of you.

Seven questions, and what a good answer sounds like

1. "Is this stone treated, and how?" Ask it first, ask it plainly, and ask for the answer in writing. A trustworthy answer is one short sentence — heated, unheated, fracture-filled, whatever it is. Vagueness is the warning sign, not treatment. Every treatment we identify goes into our treatment register, with a plain-language entry for each term in the gem lexicon.

2. "What is the origin claim based on — a laboratory report, or the seller's belief?" Both are legitimate answers. Only one of them is evidence. If it is a report, ask which laboratory and what date, and remember from Chapter III that the report itself is an opinion and that criteria change over time.

3. "Who did you buy it from, and when?" This is the provenance question rather than the origin question, and it is the one most sellers have never been asked. You are not entitled to their commercial relationships in detail, but you are entitled to know whether a record exists at all, and how far back it runs.

4. "What don't you know about this stone?" My favourite question, and the most diagnostic. No seller knows everything about a coloured stone's journey. A seller who claims to does not understand the trade they are in. Listen for whether the gap is described comfortably or defended.

5. "Is that word a claim or a certification?" Applies to "Ceylon", "conflict-free", "ethically sourced", "mine to market" and every similar phrase. Recall that African sapphire is frequently re-badged as Ceylon material, that there is no Kimberley certificate for a coloured stone, and that the ACCC's guidance specifically targets broad, unqualified claims.

6. "How do you handle AUSTRAC obligations?" An odd-sounding question that works beautifully. Since 1 July 2026, dealers who take physical currency or virtual assets of A$10,000 or more — single or linked — for precious stones have been providing a designated service and must be enrolled with AUSTRAC. Card and bank-transfer-only sellers are outside it, so the right answer may legitimately be "that limb doesn't apply to us, and here's why". What you are listening for is whether they know which side of the line they are on. A blank look in August 2026 is the warning sign, not the absence of an enrolment number.

7. "What happens if I want this re-checked in ten years?" Provenance is a long-term relationship or it is a receipt. Ask what the seller will still do for you after the sale, and get the answer in writing while they still want your business.

What to keep, and for how long

Keep everything, digitally and somewhere your executor will find it: the invoice naming the seller and the date; the written treatment disclosure; any laboratory report, with its number and issue date; photographs of the stone loose if you ever get the chance; the measured weight and dimensions; and a note of anything you were told verbally, written down while you still remember it.

This matters more than it seems, because you are not only keeping records for yourself. Gemstones are effectively imperishable and rarely have a single final owner, so the file you keep is the only thing that will travel with the stone through the estate that eventually disperses it. Records also age into evidence: a laboratory report issued before a stone changed hands again is worth considerably more than one commissioned afterwards.

Our own answer to this is the Digital Gemstone Passport, which is simply this file kept properly and kept permanently: identification, measured properties, treatment status, what we know of the stone's history, and what we do not. It travels with the stone, not with the owner.

What we can and cannot tell you about our own stones

What we can tell you. Every stone is examined in-house by an FGAA-qualified gemmologist — me — before it is listed. Treatment status is disclosed in writing on every stone, and where a stone is heated we say heated rather than reaching for a softer word. We sell only natural, earth-mined stones. Where a stone carries a laboratory report we name the laboratory and the date, and where we know who we bought a stone from and when, that goes in the Passport. Our Ceylon sapphires come through relationships my family has held in those gem fields across three generations since the 1960s; our Australian sapphires and opals come from the fields and statutory frameworks described in Chapter VI. Every stone can be brought back to us for re-authentication, free, for as long as we are here.

What we cannot tell you, and will not pretend to. For most stones we cannot name the individual pit, the licence number or the miner — for the structural reasons set out in Chapter II, and because the DNA-tagging technology that could close that gap does not survive the heat treatment most corundum receives. Where a stone is unheated it has usually skipped the treatment leg of the chain, which sometimes means one fewer set of hands to account for — but the premium on unheated stones is a rarity premium, not a provenance premium, and I would not have you pay for the two as though they were one thing. We cannot certify labour conditions at a site we have not audited. And where an origin opinion is inconclusive, we will print the word inconclusive.

On your rights: for consumer purchases, the consumer guarantees under the Australian Consumer Law apply to what we sell you and cannot be contracted away. They are a matter of law rather than of favour, and they sit alongside — not instead of — the disclosures we make voluntarily. Nothing in this handbook is legal advice, and nothing in it is investment or financial advice; where I have discussed values or premiums it has been to help you judge a price in front of you today, not to suggest what any stone will be worth later.

To our knowledge, no coloured-stone seller can currently offer complete mine-to-finger traceability across a general inventory, and I would treat any claim to the contrary as the most interesting sentence on the page — the one to ask about first.

That is the whole of it. Ask the seven questions. Keep the file. Prefer the seller who tells you what they do not know. And if you would like to test all of that on a specific stone, bring it to me — or talk to our gemmologist and we will go through it together, whether or not you bought it here.