A quantum computer decrypts standard 2048-bit RSA encryption before 2035.

Question

Six frontier AI models put this at 13% by 28 Sept 2036. Full reasoning from every model.

WillHappen.ai · Accepted Answer

13% — the median of 6 frontier models, resolving by 2036-09-28.

Answer

Gemini (Google): 14% — Breaking RSA-2048 requires millions of physical qubits with error correction, far outstripping current hardware roadmaps before 2035. Shor's algorithm requires roughly 20 million noisy physical qubits or 4,000 logical qubits; state-of-the-art remains under 2,000. Roadmaps from IBM and Google target only a few hundred logical qubits by 2030, leaving a massive scaling gap to close by 2035. A radical algorithmic breakthrough cutting needed logical qubits below 500 or sudden fault-tolerant scaling would shift odds sharply.

Answer

Claude (Anthropic): 6% — Requires millions of physical qubits with low error rates; current hardware is off by orders of magnitude. Current best quantum computers have ~1000 noisy qubits; breaking RSA-2048 needs ~20 million physical qubits with error correction per leading estimates. Error correction overhead and fault-tolerance milestones are progressing slowly, with most experts projecting cryptographically relevant quantum computers post-2035 if… NIST and industry are migrating to post-quantum cryptography as a hedge, reflecting expert consensus that RSA-2048 break by 2035 is unlikely but not impossible with a…

Answer

GPT (OpenAI): 18% — It turns on whether fault-tolerant machines reach millions of high-quality physical qubits, not today’s noisy-qubit scaling curves. Shor factoring RSA-2048 needs roughly millions of physical qubits plus long-lived error correction, far beyond 2026 systems. IBM, Google and Quantinuum must demonstrate scalable logical qubits and decoding by the early 2030s; roadmaps are not deployments. A credible million-qubit fault-tolerant prototype by 2030 would sharply raise the odds; repeated logical-error stalls would lower them.

Answer

Grok (xAI): 18% — Needs million-scale fault-tolerant qubits running Shor’s on 2048-bit RSA by 2034. IBM/Google/Quantinuum must deliver ~10k logical qubits with <10^{-12} error by ~2032; current devices are ~1k noisy physical qubits. No quantum factoring beyond RSA-15 exists; expert medians sit in the 2040s, so 9-year window sits well below base rate. A 100-logical-qubit high-fidelity demo before 2029 would raise this sharply.

Answer

DeepSeek (DeepSeek): 12% — This turns on whether a fault-tolerant machine with millions of physical qubits exists before 2035, not on qubit-count headlines. Shor's algorithm needs roughly 4,000 logical qubits and ~10^9 Toffoli gates for RSA-2048; today's devices have zero logical qubits at that scale. Google, IBM, Quantinuum and PsiQuantum roadmaps target fault tolerance in the early-to-mid 2030s, leaving little margin for a 2048-bit run by 2035. Base rate for cryptographic-scale milestones slipping: RSA-1024 factoring was 'imminent' for two decades; this case sits below even that, since error correction is the…

Answer

Qwen (Alibaba): 2% — Requires millions of physical qubits with error correction; current best machines have ~1,000 noisy qubits. Shor's algorithm for 2048-bit RSA needs ~4,000 logical qubits, implying millions of physical qubits at current error rates; IBM's roadmap targets ~100K physical qubits… No architecture has demonstrated fault-tolerant logical qubits at scale; even optimistic roadmaps (Google, IBM, IonQ) place cryptographically relevant machines in the… A breakthrough in error correction or a radically new algorithm would be needed to close the gap; neither has appeared in peer-reviewed literature as of late 2026.

A quantum computer decrypts standard 2048-bit RSA encryption before 2035.

Ethiopian pro-government forces maintain control of the newly captured strategic town in Tigray.

Russia conducts a military strike beyond Ukraine's borders against a European nation.

The United States reinstates a military draft before December 31, 2028.

A major cyberattack disables critical infrastructure in a G7 nation before December 31, 2026.