Will the prime factors of RSA-2048 be known before 2040? A long-dormant question on the prediction platform Manifold has turned into one of this week’s busiest trades, sitting at a 50% implied probability of YES as of October 18, 2026. The spike follows Google Quantum AI’s October 15 announcement that it cut a key class of quantum errors roughly 40-fold on its Willow processor, a component-level advance that has traders re-examining how close a cryptographically relevant quantum computer really is.
Don't miss new tech stories on Google
Add Tech Insider once in the Google app and our stories appear in your news suggestions.
What this prediction market is actually asking
The contract, created by forecaster Isaac King on Manifold, resolves YES if the two prime factors of the RSA-2048 modulus become publicly known before the year 2040. RSA-2048 is a 2048-bit number published as part of the old RSA Factoring Challenge; its factors have never been found. Recovering them is the canonical benchmark for “breaking” RSA-2048, the public-key scheme that still protects large parts of web traffic, software signing, and financial messaging.
A market like this is a wager on engineering timelines rather than on any single event. Classical factoring is not expected to reach a 2048-bit key. The realistic path to a YES resolution runs through a fault-tolerant quantum computer capable of executing Shor’s algorithm at scale, which is exactly why quantum hardware news moves the price. For readers new to the format, our explainer on how prediction markets work covers the mechanics of these contracts.
The odds as of October 18, 2026, and what they imply
As of October 18, 2026, the Manifold market shows a 50% implied probability that RSA-2048’s factors will be known before 2040. On the same day the market saw unusually heavy activity: roughly 26% of its all-time volume traded in a single session, with about 131 unique accounts placing positions. That concentration of flow is what flags a market as “hot,” because it signals that new information, not routine churn, is driving the repricing.
A 50% reading is best read as genuine disagreement rather than consensus. Half the weighted money thinks a working machine or a published factorization arrives inside roughly 13 years; the other half does not. For a primer on turning a percentage into a forecast, see our note on implied probability. On a play-money venue such as Manifold, the number reflects crowd conviction, not regulated capital at risk.
What is driving the move: Google’s magic state result
The catalyst is a result Google Quantum AI published in the journal Science and announced around October 15 to 16, 2026. The team demonstrated a technique called magic state cultivation on the 105-qubit Willow processor, producing so-called T-states with a reported logical fidelity of about 99.99% and roughly a 40-fold reduction in magic-state errors versus uncultivated states. Magic states are the hard-to-produce ingredient that lets an error-corrected machine run the non-Clifford operations Shor’s algorithm depends on, so cleaner, cheaper magic states are directly relevant to any future factoring engine.
Two caveats are central. First, the experiment prepared quantum states; it did not factor RSA-2048 or anything close to it. Second, a 99.99% fidelity corresponds to roughly one error per 10,000 operations, which commentators note is still several orders of magnitude away from the far cleaner operations a full RSA-2048 factoring run would require. In other words, the news improves a building block, it does not deliver the building.
From 20 million qubits to under 100,000: why estimates keep falling
The deeper reason traders are paying attention is that the paper resource estimates for breaking RSA-2048 have dropped sharply over the past several years, entirely on paper rather than in hardware. Google researcher Craig Gidney’s May 2025 analysis argued that a fault-tolerant machine with under one million physical qubits (about 897,864 in the design, with roughly 1,409 logical qubits at peak) could factor a 2048-bit key in under a week. That is about a 20-fold reduction from Gidney’s own 2019 estimate of roughly 20 million qubits over about eight hours.
The trend continued into 2026. In February 2026, Iceberg Quantum published a resource estimate claiming RSA-2048 factoring with fewer than 100,000 physical qubits by using quantum low-density parity-check (qLDPC) codes instead of surface codes. Separate neutral-atom work modeled factoring on a half-million-qubit modular atomic processor. These are simulations and architectural claims, not operational machines, but the direction of travel is what the market is pricing.
| Year | Source / design | Physical qubits (approx.) | Estimated runtime |
|---|---|---|---|
| 2012 era | Early surface-code estimates | ~1 billion | Impractical |
| 2019 | Gidney and Ekera | ~20 million | ~8 hours |
| 2025 | Gidney (Google) | <1 million | <1 week |
| Feb 2026 | Iceberg Quantum (qLDPC) | <100,000 | ~1 month |
| 2026 | Neutral-atom modular design | ~500,000 | Modeled, hundreds of days to days depending on config |
Figures are published engineering estimates and modeled architectures, not demonstrated results. Each assumes error correction, connectivity, and sustained operation that no current device delivers.
Who and what is involved
The Willow work came from Google Quantum AI, with Craig Gidney, Noah Shutty, and Cody Jones among the credited researchers. Research scientist Emma Rosenfeld described how equipment drift and temperature variation interfered with benchmarking, prompting new noise-reduction controls. On the security side, Google has set an internal goal to migrate to post-quantum cryptography by 2029.
Industry observers offered measured takes. Nic Carter of Castle Island Ventures said “there’s no imminent threat, and there remain engineering challenges to building a scaled QC,” while adding that AI is “meaningfully accelerating the rate of progress.” Alex Pruden of Project Eleven framed the result as validating existing Q-day timelines and making them “more certain” rather than pulling them forward. Those are the kinds of named voices a prediction crowd weighs when repricing a long-dated contract.
The bearish case: what has not happened
For balance, several facts cut against a dramatic repricing. No one has publicly factored an RSA-2048 modulus. No deployed RSA-2048 system has been compromised by anything reported in the past two weeks. Google’s magic state result is a component advance on a 105-qubit chip, not a scale-up toward the hundreds of thousands or millions of qubits the resource estimates assume. The sub-100,000-qubit and neutral-atom figures live in preprints and architecture studies, where assumptions about connectivity, measurement speed, reaction latency, and manufacturing yield do heavy lifting.
The honest summary is the one the reporting supports: RSA-2048 remains unbroken, the recent catalyst is better error-correction and architecture estimates rather than a factoring breakthrough, and “before 2040” is technically plausible but highly assumption-dependent.
Broader context for a tech and finance reader
The money at stake here is small and play-money, but the underlying risk is not. If a cryptographically relevant quantum computer arrives this decade or next, it would threaten the confidentiality of data already harvested under “store now, decrypt later” strategies, as well as the integrity of code-signing and financial messaging that still lean on RSA. That is why the US National Institute of Standards and Technology finalized its first post-quantum cryptography standards in 2024 and why large technology firms are setting internal migration deadlines.
Prediction markets have become a quick sentiment gauge for exactly these hard-to-date risks, the same way traders watch Fed rate-cut odds for macro expectations. Manifold’s play-money model differs from cash venues; our comparison of Kalshi versus Polymarket explains how real-money exchanges price similar questions, and this contract sits alongside other offbeat Manifold bets such as a recent viral personal-life market.
What to watch next
Three signals would matter more than headlines. First, any demonstration of magic state cultivation at larger scale, which Google itself says is the next step. Second, peer-reviewed resource estimates that survive scrutiny rather than preprints, especially designs claiming sub-100,000 qubits. Third, concrete hardware roadmaps showing logical-qubit counts climbing toward the four-figure range Shor’s algorithm needs. Absent those, expect the market to drift on sentiment rather than fundamentals, and expect volume spikes to follow each new quantum paper.
Frequently asked questions
Has RSA-2048 been broken? No. As of October 18, 2026, no one has publicly factored the RSA-2048 modulus, and no deployed RSA-2048 system has been compromised by the recent quantum results.
What did Google actually announce? Google Quantum AI demonstrated magic state cultivation on its Willow processor, reporting about 99.99% logical fidelity and roughly a 40-fold reduction in magic-state errors. It prepared quantum states and did not run a factoring algorithm.
Why did the Manifold market move? Traders read the Google result, together with falling paper estimates for the qubits needed to factor RSA-2048, as modest support for a pre-2040 break. The contract traded about 26% of its all-time volume on October 18, 2026.
Is a 50% probability a prediction that it will happen? No. It reflects roughly even disagreement among participants on a play-money venue, not a calibrated forecast or a regulated-capital signal.
The Bottom Line
Sources
- The Block — Google demonstrates quantum computing breakthrough as observers weigh Q-day implications
- StreetInsider — Google demonstrates magic state cultivation on Willow quantum processor
- Quantum Zeitgeist — Google cuts quantum error rates by 40x with magic state cultivation
- The Quantum Insider — New architecture could cut quantum hardware needed to break RSA-2048 tenfold
- PostQuantum — Quantum breakthrough slashes qubit needs for RSA-2048 factoring
- Manifold — Will the prime factors of RSA-2048 be known before 2040?
Prediction markets carry financial risk and are not investment, legal, or betting advice. Market availability is restricted by jurisdiction: Polymarket is not available to US persons, while Kalshi is a CFTC-regulated US exchange, and Manifold operates on play-money with no cash payouts. Participation is limited to those of legal age (18+ or 21+ as applicable in your location). If gambling is affecting you or someone you know, call 1-800-GAMBLER for confidential help.
Related Coverage
- Yang-Mills "Solved Next Before 2045?" Odds Jump 20 Points on Kalshi Amid an AI Math Frenzy
- Tulane CFP Odds Swing +35 Points on Kalshi: What Moved the Market
- Taiwan Level 4 Travel Warning Odds Jump to 69% on Kalshi After China's Multi-Sea Drills
- GiveWell Cash Transfer Grant: Why Manifold Traders Just Bet It to 99% Yes
- Perplexity vs ChatGPT vs Google AI Mode: 37% vs 67% [2026]
