Lab-Grown Gold: Could It Crash Gold Prices?

A scientific breakthrough is only the beginning

If laboratories can manufacture gold, what happens to the jewellery, coins and investments people already own? The question has returned with Marathon Fusion’s claim that it has converted mercury into gold, discussed in Civil Beings’ video published on 28 October 2026.

The concern makes sense: a technology that eventually delivers abundant, inexpensive gold could change its economics. But several separate achievements lie between creating gold atoms and supplying a competitive product. A laboratory result, a model of a future reactor and a profitable production line each answer a different question.

The useful starting point is to examine the science, the quantities and the conditions required for commercial supply. A headline alone cannot tell us how much gold will reach buyers—or when.

How mercury can become real gold

An element’s identity comes from the number of protons in its nucleus. Gold has 79; mercury has 80. Ordinary chemical processing cannot make that change. Nuclear transmutation can.

In Marathon’s proposed pathway, a fast neutron strikes mercury-198, leaving mercury-197 after two neutrons emerge. Mercury-197 then undergoes electron capture: a proton becomes a neutron, producing stable gold-197. Its ground-state half-life is about 64 hours. That describes the decay of mercury-197 already produced—not the conversion of half an entire mercury supply every few days. [1]

The result is genuine gold at the atomic level. Whether a recovered batch is sufficiently pure and suitable for sale is a separate question.

What CERN’s tiny quantity actually tells us

CERN reported that the Large Hadron Collider’s four major experiments produced about 86 billion gold nuclei during Run 2, from 2015 to 2018. Their combined mass was just 29 picograms: 29 trillionths of a gram. These nuclei rapidly fragmented after striking accelerator components; they were not collected as usable bullion. [2]

CERN’s lead-to-gold process also differs from Marathon’s mercury route. A claim of beating a production record by ten times therefore needs a clearly defined comparison: the mass produced, the rate, the operating time and the amount recovered. Multiplying a microscopic starting quantity does not, by itself, establish an industrial supply.

A reactor forecast is not a delivery of bullion

Marathon’s research models approximately two tonnes of gold annually per gigawatt of thermal fusion power. This is a simulation-based estimate. Thermal power measures heat; electrical output depends on conversion efficiency, so the two capacities cannot be substituted for each other. [1]

The company proposes using gold as a valuable co-product while the plant also produces energy and replenishes its fusion fuel. [3] That would be an attractive business model if it works at the intended scale. It would still require operating equipment, reliable production and customers able to accept the product.

Think of the difference between designing a factory and reporting its audited annual deliveries. The design helps explain the opportunity. The deliveries demonstrate how much supply the market can actually receive.

Would another 500 tonnes change the market?

Consider the video’s hypothetical scenario of 100 plants producing five tonnes each: 500 tonnes a year. Treat this as an illustration, not a forecast of plants being built.

The World Gold Council’s January 2026 report estimated 2025 mine production at 3,671.6 tonnes and total supply at 5,002.3 tonnes, including recycling and producer hedging. Its separate stock estimate put gold above ground at approximately 222,600 tonnes at the end of June 2026. [4, 5]

On those figures, 500 tonnes would equal about 13.6% of annual mine production, 10% of annual total supply and 0.22% of the accumulated stock. These are our calculations. The comparison matters: a small addition to historical holdings can still be a substantial change in yearly supply.

There is no direct formula turning those percentages into a price fall. The outcome would also depend on demand, existing holders’ willingness to sell, production costs and expectations about further expansion. The scenario demonstrates why scale deserves attention; it does not establish an imminent crash.

Why the diamond comparison has limits

Laboratory-grown diamonds are real diamonds. GIA explains that they share essentially the same chemical, physical and optical properties as natural stones, although specialised examination can distinguish their origin. They are made through processes such as high pressure and temperature or chemical vapour deposition. [6]

Making a diamond arranges carbon into a crystal. Making gold from another element changes atomic nuclei. Success in scaling one process does not establish the cost or timetable for the other.

Gold also serves several markets. The World Gold Council’s stock breakdown includes substantial holdings in jewellery, central-bank reserves, bars, coins and gold ETFs. [5] Those different buyers make a simple copy of the diamond-price story unreliable. Technology could change gold’s scarcity premium, but a prediction that gold must become worthless by a particular year goes beyond the evidence.

Four developments that would change the story

First, independently checked output. Look for repeatable measurements of recovered gold, including quantity and purity. A dramatic comparison with an earlier record is less informative than results that other specialists can evaluate.

Second, sustained production at a meaningful scale. An operating system needs to demonstrate dependable output over time. The important question becomes how much additional gold can consistently reach the market.

Third, credible costs. A useful business case must account for the plant, feedstock preparation, processing, maintenance, financing and any delay before sale. Calling gold a byproduct does not make those costs disappear.

Fourth, a product suitable for commercial use. Marathon’s paper discusses residual radioactivity and potentially lengthy storage; the timing depends on the isotope mixture and purification. [1] Creating stable gold atoms and delivering an acceptable batch of metal are different milestones.

What gold holders can take from this

The central lesson is to follow the evidence between discovery and deployment. The reported experiment and the reactor projections do not establish a near-term flood of inexpensive, saleable gold. They also provide no guarantee that technological progress will leave the market unchanged forever.

For someone evaluating a gold holding, this development is one factor to monitor alongside the purpose of the holding, diversification, liquidity needs and time horizon. It does not supply a universal instruction to buy, sell or hold.

The question worth returning to is concrete: can a producer repeatedly deliver meaningful quantities of acceptable gold at a competitive cost? Progress on that question will tell us far more than the headline.

Sources and editorial notes

Prepared from the supplied Civil Beings video transcript, with scientific units and market figures checked against the sources below. Context and research checked on 28 October 2026. The 500-tonne example is hypothetical; percentages are calculated from the stated World Gold Council figures.

[1] Marathon Fusion, Scalable Chrysopoeia via (n, 2n) Reactions Driven by Deuterium-Tritium Fusion Neutrons, research paper (2025). https://www.marathonfusion.com/alchemy.pdf

[2] CERN, ALICE detects the conversion of lead into gold at the LHC (2025). https://home.cern/alice-detects-conversion-lead-gold-lhc/

[3] Marathon Fusion, Alchemy: Growing fusion’s fuel supply while co-producing gold. https://www.marathonfusion.com/alchemy/

[4] World Gold Council, Gold Demand Trends: Q4 and Full Year 2025 — Supply, 29 January 2026. https://www.gold.org/goldhub/research/gold-demand-trends/gold-demand-trends-full-year-2025/supply

[5] World Gold Council, Above-ground stock, 18 August 2026; figures through the end of Q2 2026. https://www.gold.org/goldhub/data/how-much-gold

[6] GIA, Is There a Difference Between Natural and Laboratory-Grown Diamonds?, 31 May 2019. https://www.gia.edu/gia-news-research/difference-between-natural-laboratory-grown-diamonds