The Soviet GKZ system (A/B/C1/C2 explored reserves, P1/P2/P3 prognostic resources) classifies deposits by exploration density alone, with no requirement for economic viability. JORC, NI 43-101, and other CRIRSCO-family codes require Modifying Factors and a Competent Person's sign-off. No GKZ category maps one-to-one onto Measured, Indicated, or Inferred; conversion requires re-estimation, not relabeling.
GKZ is the reserves and resource classification built by the State Commission on Reserves of the USSR (Gosudarstvennaya Komissiya po Zapasam) and still in use, with local variants, across most of the former Soviet space. It splits deposits into two families rather than one continuous scale. Explored reserves use categories A, B, C1, and C2, ordered from densest drilling and highest confidence (A) down to the minimum grid needed to prove continuity (C2). Prognostic resources use P1, P2, and P3, which are forecasts of what a region or a specific structure might hold based on geological analogy, not delineated by drilling at all. A deposit report under GKZ typically states balance reserves (economically extractable under the state-approved cut-off at the time of approval) and off-balance reserves (below that cut-off but geologically real) separately for each category.
The practical difference from CRIRSCO-family codes shows up immediately: GKZ has seven tiers describing geological confidence, while JORC has three Mineral Resource tiers (Measured, Indicated, Inferred) plus two Ore Reserve tiers (Proved, Probable) that layer economic and technical confirmation on top of geology. The tier counts do not line up, and neither does what each tier is trying to say.
GKZ was built for a single client: the state. Every deposit in the USSR belonged to the state, every mine plan was approved centrally, and the reserve report existed to feed a planning apparatus, not to inform a buy or sell decision by an outside investor. Category assignment ran through regional geological departments and expert commissions, approved by protocol, with no single named person carrying legal liability for the number. Cut-off grades were fixed by state economic planning and were, in practice, rarely revisited between approval cycles, because the "market" that would normally force a revision did not exist.
That design is coherent for its purpose. It breaks down the moment a report has to answer a different question: would a rational private investor, using today's costs and today's prices, extract this material at a profit. GKZ was never built to answer that question, and re-reading a GKZ report as if it already contains the answer is the single most common error in cross-border technical reporting on former Soviet assets.
The direct answer: GKZ measures how well a deposit has been geologically delineated; JORC, and every other CRIRSCO-family code, measures whether that same material can be reported as a Mineral Resource or Ore Reserve only after Modifying Factors (mining, metallurgical, economic, marketing, legal, environmental, social, and governmental factors) have been applied by a named Competent Person who takes personal responsibility for the estimate. GKZ counts what is there. JORC reports what can be mined at a profit, signed by someone who can be held accountable for saying so.
Three structural gaps follow from that one difference. First, accountability: GKZ approval is collective, through a commission and a protocol; JORC and its sister codes require one identifiable Competent Person, a member of a recognised professional body, whose name is attached to the public statement. Second, boundary logic: GKZ reserves are typically bounded by a hard economic cut-off grade fixed at approval, excluding everything below it outright, while CRIRSCO codes allow a geological mineralisation envelope with internal dilution, then apply the economic cut-off separately during Reserve conversion. Third, currency: a GKZ approval can sit unrevised for years, while a JORC-style Mineral Resource statement is expected to be updated within weeks of a material change (new drilling, a metal price move, a permitting event). None of these three gaps is fixable by relabelling a spreadsheet column.
CRIRSCO's own guidance on this conversion is explicit that it runs one way: from Russian-family classifications toward the international template, not back. The Russian system has more categories than the CRIRSCO template has boxes to put them in, so a reverse conversion cannot be made to work symmetrically.
The direct answer: there is no safe default mapping you can apply without knowing the deposit. The same GKZ category can land in different CRIRSCO tiers depending on deposit complexity class, whether the estimate went through formal GKZ registration and expert review, and how much of the Modifying Factors work still needs to be done from scratch.
Two variables drive most of the divergence. Registration status: an "author's estimate" that was never submitted to GKZ for registration has not passed independent expert review, and the working guidance from the joint CRIRSCO-GKZ effort is that such unregistered C1 or C2 material should not be converted higher than Inferred, regardless of how it reads on paper. Deposit complexity: GKZ assigns deposits a complexity class from 1 (simple, tabular, continuous) to 4 (structurally complex, discontinuous); in a complexity-4 deposit there is typically little true C1 material, and what exists usually maps down to Indicated rather than Measured because the drilling density that would justify Measured confidence for that geometry is higher than what C1 required under GKZ.
The direct answer: registered A and B material is the strongest candidate for Measured or Indicated Mineral Resources, registered C1 splits between Measured and Indicated depending on complexity class, C2 splits between Indicated and Inferred, and P1 through P3 generally do not qualify as Mineral Resources at all under CRIRSCO codes; at best P1 supports an Exploration Target statement. Every cell in the table below still requires the receiving Competent Person's own verification. Treat it as a starting hypothesis, not a lookup table.
| GKZ category | What it represents | Typical CRIRSCO-family landing zone | Conditions that change the outcome |
|---|---|---|---|
| A | Highest-density drilling, deposit geometry and grade continuity fully proven | Measured Mineral Resource (candidate) | Requires Modifying Factors and current economic cut-off; simple-geometry deposits only |
| B | Dense drilling, minor extrapolation between points | Measured or Indicated Mineral Resource | Complexity class and grade continuity between points decide which tier |
| C1 | Moderate drilling density, continuity inferred between wider-spaced points | Indicated Mineral Resource, sometimes Measured | GKZ registration status; complexity class 4 deposits usually cap at Indicated |
| C2 | Sparse drilling, continuity extrapolated from geological analogy plus limited data | Inferred Mineral Resource, sometimes Indicated | Unregistered author's estimates should not exceed Inferred |
| P1 | Forecast resources adjacent to a known deposit, geological extrapolation only | Usually no Mineral Resource status; at best supports an Exploration Target | No drilling-based continuity; not a JORC resource category |
| P2 / P3 | Regional prognostic potential based on analogy and geophysical inference | No CRIRSCO equivalent | Reportable only as background geological potential, never as a Resource or Reserve figure |
Read the right-hand column as the actual instruction. The category letter tells you how the Soviet system rated the drilling; it does not tell you what a Competent Person can defensibly sign today.
The direct answer: a Competent Person converting a GKZ report has to re-verify the underlying data, apply Modifying Factors that GKZ never required, and re-run the resource estimate under a current economic cut-off, not just re-stamp the old category letters with new English names. In practice that means five separate workstreams. Data verification: original drill logs, assay certificates, and survey records need to be checked against the summary tables in the GKZ protocol, because Soviet-era reports commonly carry forward rounding and transcription choices that were fine for a state accounting exercise but fail an audit trail test. Coordinate control: collar and section coordinates in Soviet archives are almost always in Pulkovo 1942, which has to be reprojected to WGS-84 (or the relevant local datum) before the drillhole geometry can be loaded into a modern resource model. Cut-off re-estimation: the GKZ cut-off was fixed at approval and often years or decades out of date; a current cut-off requires today's costs, prices, and recoveries. Modifying Factors: mining method, metallurgical recovery, permitting status, and market access all have to be assessed from scratch, since GKZ registration only partially covers this ground. Independent QA/QC: Soviet-era assay QA/QC rarely meets modern check-sample and standard-insertion protocols, so twin drilling or check assaying is frequently needed before historical grades can be relied on at all. The methodology a Competent Person applies to this work is the same one used for any historical-data resource estimate; GKZ archives are simply a specific and well-documented case of it.
The direct answer: conversion (or an equivalent CRIRSCO-aligned code) is mandatory wherever a company reports mineral assets to a securities regulator or stock exchange that has adopted a CRIRSCO-family code, and it is increasingly required inside Kazakhstan itself under KAZRC. ASX-listed companies must report under JORC; TSX and TSX-V listed companies must report under NI 43-101, which itself defers to CRIRSCO-aligned definitions; LSE-listed and AIM companies report under the code the company nominates, almost always a CRIRSCO-family code; US SEC registrants have reported under Regulation S-K Subpart 1300 since it became mandatory for fiscal years beginning on or after 1 January 2021, replacing the old Industry Guide 7 and explicitly built on CRIRSCO principles. Inside Kazakhstan, the Kazakhstan Reporting Code (KAZRC) became the country's CRIRSCO-recognised national standard after Kazakhstan joined CRIRSCO as its tenth member in 2016, and the government has been moving new deposits onto CRIRSCO-aligned reporting rather than legacy GKZ approval alone. A company sitting on GKZ-only reserve numbers cannot use those figures directly in an ASX prospectus, a TSX technical report, an SEC filing, or, increasingly, a domestic KAZRC-compliant filing; the conversion work described above is a prerequisite for capital-raising, not a formatting exercise.
The direct answer: the parts of this problem that are mechanical (reading typewritten Cyrillic reports, reprojecting coordinates, and organising category data into a structure a Competent Person can audit) are exactly the parts automation should carry, so the geologist's time goes to the judgment calls that cannot be automated: complexity classification, Modifying Factors, and the final sign-off. Three tasks dominate the mechanical load. OCR on typewritten and often carbon-copied Soviet-era protocols, tables, and maps, most of which were never digitised and exist only as scanned or physical archive pages. Coordinate reprojection from Pulkovo 1942 (and related local Gauss-Kruger zones) to WGS-84, applied consistently across every drillhole and section in a deposit file rather than by hand, point by point. Category-to-category mapping support that surfaces the deposit's complexity class, registration status, and drilling density automatically, so the Competent Person starts from a structured hypothesis (per the table above) instead of re-reading hundreds of pages of protocol text to find the same facts. A platform such as CoreElement AI's Soviet GKZ conversion module packages exactly this pipeline; see the practical walkthrough in the step-by-step GKZ to JORC conversion guide for how the individual steps chain together on a real archive, and the Soviet archive digitisation page for the digitisation side specifically.
CoreElement AI was founded in 2024 by Daniel Tonkopiy (CEO and Product Architect) and Zhambyl Suraganov (Cofounder). Its platform runs 22 modules across the reporting codes covered on this page (JORC 2012, NI 43-101, KAZRC, SAMREC, PERC 2021, S-K 1300, with CRIRSCO mapping throughout), and its Soviet GKZ A/B/C1/C2/P conversion module combines OCR of archive documents, Pulkovo 1942 to WGS-84 reprojection, and category mapping support of the kind described above. As of May 2026 the module has processed 4,859 digitised documents, and the underlying drill-hit ranking model has shown a 76% AI-ranked drill hit rate in back-testing. The module has run in two pilot deployments with large enterprise mining operators in Kazakhstan. It does not replace the Competent Person's sign-off; it removes the archive-reading and coordinate-transformation bottleneck that otherwise consumes most of the time on a Soviet-era conversion project.