Forecast
Industrial resource extraction from the Moon or asteroids operates by 2126.
Last revised Sep 10, 2026.
Current estimate
probability this has happened by each date
Extraction of material from the Moon or an asteroid operates for at least twelve continuous months, supplying material that is sold or consumed by an actual operation, documented by the operating agency or company. A single demonstration payload, a sample return, and a plant that produces without its output being used all fail the test.
Resolution criteria
- 2126
- Extraction of material from the Moon or an asteroid operates for at least twelve continuous months, supplying material that is sold or consumed by an actual operation, documented by the operating agency or company. A single demonstration payload, a sample return, and a plant that produces without its output being used all fail the test.
Probability history
| Revised | State | p(2035) | p(2050) | p(2100) |
|---|---|---|---|---|
hf-forecast-0019 | researched | — | — | — |
hf-forecast-0012 | initial prior | — | — | — |
Strongest evidence for
- Launch cost fell from 87,023 to 3,868 dollars per kilogram between 1960 and 2025, at 21.2 percent per doubling of cumulative payload with an R-squared of 0.953 since 2000 (hf-evidence-0020).
- The resource exists: water ice is directly observed in permanently shadowed lunar polar regions within the top few millimetres of the surface (hf-evidence-0022).
- In-situ production is not hypothetical. MOXIE made 122 grams of oxygen from the Martian atmosphere across 16 runs at 98 percent purity (hf-evidence-0018).
- The criterion admits operational use as well as sale, so a sustained lunar presence consuming its own oxygen or propellant would resolve it without any market existing at all.
- Every obstacle found by an adversarial search is a maturity curve rather than a wall, and the adversary said so explicitly: none of the counterevidence supports not-this-century.
Strongest evidence against
- The total demonstrated off-Earth material haul across all missions ever is on the order of kilograms: 122 grams of oxygen made on Mars, 121.6 grams returned from Bennu, and roughly 1.9 kilograms from the lunar far side (hf-evidence-0018, hf-evidence-0021).
- The only flight carrying lunar resource-extraction hardware failed: IM-2 landed on its side and its mass spectrometer detected its own lander's propulsion gases rather than lunar water (hf-evidence-0019).
- Commercial lunar landing is not yet reliable, with one fully successful landing in seven private attempts from 2019 to 2025 (hf-evidence-0026).
- Both companies founded to mine asteroids died without extracting anything (hf-evidence-0023).
- No buyer exists outside the extraction industry itself, and PwC's own bull case has transport consuming 70 to 80 percent of lunar infrastructure cost through 2035 (hf-evidence-0024).
- The programmes that would create a sustained presence slip: GAO found Artemis III unlikely to meet December 2025, and it moved to September 2026 (hf-evidence-0025).
- Title to extracted resources remains contested under Outer Space Treaty Article II (hf-evidence-0027).
Present uncertainties
- Whether anything outside the extraction industry ever buys off-Earth material, and whether a sustained presence chooses to make consumables or import them.
Next discriminating observation
The first purchase of off-Earth material by a buyer not itself in the extraction business; and the first lunar surface plant that runs for twelve continuous months with its output actually consumed.
Sources
- NASA's oxygen-generating experiment MOXIE completes Mars mission
MOXIE generated 122 grams of oxygen in total across 16 production runs aboard Perseverance between 2021 and 2023, at a peak rate of 12 grams per hour and 98 percent purity or better.
Limit: An explicit technology demonstration on Mars rather than the Moon or an asteroid, at a scale minuscule relative to any operational need. The source does not state whether the oxygen was stored, used, or vented, so no claim is made here about its fate; it was not sold or put to sustained operational use either way.
Fetched and confirmed directly by the master agent.
- NASA receives some data before Intuitive Machines ends lunar mission
The IM-2 Athena lander touched down on 6 March 2025 on its side, roughly 1,300 feet from its intended Mons Mouton target, and battery depletion ended the mission. The TRIDENT drill demonstrated its full range of motion but there is no indication it extracted regolith; MSOLO detected no lunar water, instead detecting elements likely emitted by the lander's own propulsion system. NASA stated the mission did not achieve all of its objectives.
Limit: A single mission failure caused by landing attitude, not by the extraction hardware being unworkable; the drill mechanism itself functioned. One failure does not establish that the approach cannot work.
Fetched and confirmed directly by the master agent.
- From Sputnik to Starship: estimating the experience curve of space launch technology
Across 4,405 launches from 1960 to 2025, cost to orbit fell from 87,023 US dollars per kilogram in 1960 to 3,868 in 2025. Cost falls 21.2 percent for each doubling of cumulative payload, with an R-squared of 0.791 overall and 0.953 for 2000 onwards. The central scenario, assuming payload growth continues at the 13.3 percent annual rate seen from 2010 to 2025, projects 1,600 dollars per kilogram by 2030 and 300 by 2040.
Limit: A fitted experience curve, not a guarantee, dependent on continued payload growth. The paper does not project beyond 2040, so it says nothing directly about 2126. It measures the cost of lifting mass off Earth, which is the opposite direction from extraction economics, and is silent on processing, power, equipment mass, and return logistics.
Fetched and confirmed directly by the master agent.
- NASA announces OSIRIS-REx bulk sample mass
The OSIRIS-REx mission returned 121.6 grams of material from asteroid Bennu, against a mission requirement of at least 60 grams, from a single touch-and-go sampling event on a mission that launched in 2016 and returned in 2023.
Limit: A one-time scientific return rather than sustained throughput, and grams rather than the tonnes any industrial claim implies.
Retrieved by a delegated source worker from a primary page and not independently re-fetched.
- Ice confirmed at the Moon's poles
Moon Mineralogy Mapper data provided direct observation of water ice in permanently shadowed regions at both lunar poles, concentrated in craters at the south pole and more sparsely spread at the north, sitting within the top few millimetres of the surface where temperatures never rise above about minus 157 degrees Celsius.
Limit: A spectroscopic detection of surface ice signatures and their distribution, not a mass-concentration measurement. It establishes presence and accessibility in principle, not how much ice underlies a given area. Widely cited concentration figures from the LCROSS impact could not be confirmed against primary text in this cycle and are deliberately excluded from this record.
Retrieved by a delegated source worker from a primary page and not independently re-fetched.
- The 2012 asteroid-mining cohort did not survive to attempt extraction
Both companies founded explicitly to mine asteroids ceased to exist as independent ventures without extracting anything: Planetary Resources was acquired by ConsenSys in 2018 after failing to raise its next round, and Deep Space Industries was acquired by Bradford Space in 2019 and redirected to spacecraft propulsion.
Limit: Two failures in one funding cycle over roughly seven years say little across a hundred-year horizon, and they failed commercially rather than being shown to be technically impossible. A later cohort with cheaper launch is a different experiment.
Sourced by a delegated worker from science journalism rather than a primary document; the individual mission outcomes are uncontested public record, but the tally itself was not read from a primary source.
- The demand case for space resources is currently self-referential
No private customer with a signed contract for lunar or asteroid material was identified. The stated demand case is largely missions that exist to buy propellant from missions that exist to sell it, with no buyer outside that loop. PwC's own lunar market assessment projects transportation consuming 70 to 80 percent of all lunar infrastructure cost from 2026 to 2035, falling only to 50 to 60 percent by 2046 to 2050, and treats resource use as a way to reduce the cost of other services rather than as a distinct sales market.
Limit: Analytical commentary and an interested-party consulting forecast, not peer-reviewed economics, and the PwC figures were not read from the primary document. It argues demand is investment-contingent and circular today, which is a chicken-and-egg problem rather than a permanent wall. Critically, this forecast's criterion admits operational use as well as sale, so an absent market does not by itself defeat the claim.
Delegated; the commentary page was fetched by the worker, the PwC report was not.
- NASA Artemis programs: lunar landing plans are progressing but challenges remain
GAO concluded in November 2023 that the Artemis III lunar landing was unlikely to occur in December 2025 as planned given delays and remaining technical work, and NASA adjusted the date to September 2026 to allow contractors to complete a significant amount of remaining complex work. GAO also found no established official cost estimate for the full mission and flagged ambitious schedules and acquisition management as portfolio-wide risks.
Limit: Schedule slippage in one programme over a few years is weak evidence across a century, and it reflects management and funding rather than a technical ceiling.
Fetched and confirmed directly by the master agent.
- Commercial lunar landing is not yet reliable
Of seven commercial or private lunar landing attempts from 2019 to 2025, exactly one, Firefly's Blue Ghost Mission 1 in March 2025, achieved a fully successful upright and fully operational landing. SpaceIL and both ispace missions crashed, Astrobotic's Peregrine never reached the Moon, and both Intuitive Machines landers tipped over.
Limit: A landing record, not an extraction record, over six years of an immature commercial sector. Firefly's clean success shows the capability exists and is not fundamentally unsolvable. A separate widely circulated tally of six successes in fifteen attempts including government missions could not be confirmed against a primary source and is excluded.
Sourced by a delegated worker from science journalism rather than a primary document; the individual mission outcomes are uncontested public record, but the tally itself was not read from a primary source.
- Outer Space Treaty Article II leaves title to extracted resources contested
Article II of the 1967 Outer Space Treaty provides that outer space, including the Moon and other celestial bodies, is not subject to national appropriation. Legal analysis argues that national statutes granting ownership of extracted resources, and the Artemis Accords framework built around them, do not by themselves settle whether such appropriation is lawful under the treaty.
Limit: One scholarly reading rather than a settled position or a ruling, and its exact publication date was not established. The primary Congressional Research Service analysis on this question could not be retrieved this cycle, so this record is thin and should be strengthened before it carries weight. Legal ambiguity is a political-practice problem that a century is ample time to resolve in either direction.
Delegated; primary CRS source unavailable this cycle and recorded as an open research gap.