Quantum Hacking in 2026: Could 23,000 Qubits Break RSA-2048 in Just 3 Days?

 

Quantum Hacking in 2026: Could 23,000 Qubits Break RSA-2048 in Just 3 Days?

RSA-2048 is one of the most widely used encryption standards in the world. It protects online banking, websites, social media accounts, emails, software updates, digital signatures, and stored data.

For decades, its security has rested on one simple idea: factoring a very large number is easy to set up, but extremely hard to reverse.

A new study published in 2026 suggests that a sufficiently powerful quantum computer could eventually break that lock, and it may need far fewer qubits than scientists previously believed.

This does not mean RSA-2048 has already been hacked. No existing quantum computer can do this today. But the new estimate has changed how seriously governments, banks, and tech companies are treating the quantum threat.

What Did the New Study Find?

French quantum computing company Alice & Bob, based in Paris and Boston, published a detailed resource estimate for running Shor's algorithm on its cat-qubit hardware.

Its headline results were:

• RSA-2048: 22,593 physical cat qubits in 3.14 days
• Faster RSA-2048 option: 610,530 physical cat qubits in 1.25 hours
• Slower RSA-2048 option: 21,188 physical cat qubits in 8.08 days
• ECDLP-256 / Bitcoin elliptic-curve cryptography: up to 29,588 physical qubits in 61.8 minutes

Alice & Bob says its cat-qubit architecture needs 40 times fewer physical qubits than Google Quantum AI researcher Craig Gidney's 2025 estimate for breaking RSA-2048.

The fast configuration is especially important because it shows the trade-off at the heart of quantum computing: you can use more qubits to finish faster, or fewer qubits if you are willing to wait longer.

Important correction: Some social media posts say the fast system would take 1.5 hours. Alice & Bob's official table lists 1.25 hours, not 1.5 hours.

What Is Shor's Algorithm?

Shor's algorithm was published by mathematician Peter Shor in 1994.

Classical computers struggle to factor large numbers because the time required grows enormously as the number gets bigger. A 2048-bit RSA key would take a classical computer far longer than the age of the universe to factor using known methods.

Shor's algorithm changes the math. A large, error-corrected quantum computer running Shor's algorithm could factor that same number in hours or days, not billions of years.

That is why Shor's algorithm is considered the defining benchmark for fault-tolerant quantum computing: it has a clear problem, a clear answer, and a clear quantum advantage if the hardware can be built.

It does not break all encryption equally. Shor's algorithm threatens public-key systems such as RSA and elliptic-curve cryptography. Symmetric encryption such as AES-256 is not broken by Shor's algorithm in the same way.

Why Are the Qubit Estimates Falling So Fast?

The new estimate did not come from one breakthrough alone. It came from improvements in both algorithms and hardware design.

1. Cat qubits

Alice & Bob uses superconducting cat qubits, named after Schrödinger's cat. These qubits suppress bit-flip errors at the hardware level, so error correction mainly needs to handle phase-flip errors. That sharply reduces the number of physical qubits needed per logical qubit.

2. LDPC-cat codes

The team moved from repetition codes to more efficient LDPC-cat codes. Alice & Bob says this change alone more than halved the qubit count for RSA-2048, from 49,649 to 22,593.

3. A redesigned adder

Addition is one of the operations Shor's algorithm performs most often. The researchers rebuilt the adder for cat qubits so it runs at a steady speed without repeatedly pausing for error correction.

4. Better algorithms

The team also built on recent algorithmic improvements from Inria and Google, making the overall computation more efficient.

The long-term trend is striking:

Year	Estimate to Break RSA-2048
2019	About 20 million physical qubits
2023	About 350,000 cat qubits
2024	Just under 100,000 cat qubits
2025	Under 1 million noisy qubits in Google's updated estimate
2026	About 22,593 cat qubits in 3.14 days

Alice & Bob describes this as an exponential downward trend since 2019, driven by both better algorithms and better error correction.

Can Quantum Computers Hack RSA-2048 Today?

No.

No one currently has a functioning system with 23,000 reliable, error-corrected qubits. Current quantum computers are still far too small and too error-prone to run Shor's algorithm against real-world key sizes.

The 2026 paper is a resource estimate, not a successful hack. It is also a preprint, meaning it had not yet undergone peer review when it was posted to the Cryptology ePrint Archive.

There is also an important assumption behind the headline number. The 22,593-qubit figure assumes almost no bit-flips during the computation, equivalent to a bit-flip time of about one century. Alice & Bob notes that if bit-flips occurred about once an hour instead, RSA-2048 would need about 45,000 qubits.

Alice & Bob itself says substantial engineering challenges remain, but it expects hardware improvements and falling resource requirements to converge in the early 2030s.

In other words: the threat is not here today, but the timeline is getting shorter.

What About Bitcoin and Blockchain?

The study also estimated the resources needed to break ECDLP-256, the elliptic-curve problem behind ECDSA signatures used by Bitcoin and Ethereum.

That would take up to 29,588 physical qubits and about 61.8 minutes.

That timing matters. Bitcoin's block time is about 10 minutes, so Alice & Bob notes that a transaction in progress would remain out of reach because the attack would take longer than one block. But exposed public keys, reused addresses, and long-lived wallets would face greater risk in a post-quantum future.

Earlier in 2026, Google researchers also warned that a future quantum machine could break Bitcoin wallet cryptography in less than 10 minutes under different assumptions, showing that estimates vary widely depending on hardware and algorithm assumptions.

The Bigger Threat: Harvest Now, Decrypt Later

Even if a quantum computer capable of breaking RSA-2048 does not exist for years, attackers do not need to wait.

In a “harvest now, decrypt later” attack, hackers steal encrypted data today and store it until quantum computers become powerful enough to decrypt it later.

That is why long-lived secrets are the most urgent concern: government records, medical data, trade secrets, legal files, and encrypted communications that must stay private for 10, 20, or 30 years.

The 2026 Thales Data Threat Report found that harvest now, decrypt later was the leading quantum security concern, cited by 61% of respondents.

Governments Are Already Moving

On September 3, 2026, the G7 Cyber Security Working Group and the U.S. CISA issued a joint advisory urging organizations to begin moving to post-quantum cryptography now. The advisory was signed by agencies from the USA, UK, France, Germany, Canada, Japan, and Italy.

Alice & Bob also notes that U.S. and UK guidance calls for organizations to complete the transition to post-quantum cryptography by 2035.

Separately, NIST finalized its first post-quantum replacement standards in August 2024, and its timeline plans to deprecate vulnerable algorithms after 2030 and disallow them after 2035.

The message from governments is consistent: migration takes years, so organizations cannot wait until a quantum breakthrough is announced.

What Should Organizations Do Now?

Researchers are not telling everyone to panic. They are telling organizations to prepare.

1. Discover your cryptography
Find where RSA, ECDSA, ECDH, and other vulnerable public-key cryptography are used: websites, VPNs, email, cloud services, APIs, IoT devices, and stored archives.

2. Prioritize long-lived data
Protect data that must remain secret for many years first. Short-lived session keys matter less than decades-long secrets.

3. Adopt post-quantum cryptography
Test and deploy NIST's post-quantum standards, such as ML-KEM for key exchange. Many organizations are now using hybrid systems that combine classical and post-quantum encryption during the transition.

4. Build crypto-agility
Design systems so cryptographic algorithms can be replaced quickly without rebuilding the entire product. The next breakthrough may not be the last one.

5. Plan for the 2030-2035 deadlines
Large migrations often take 5 to 10 years. Organizations that start now will have time to test, fix compatibility problems, and train teams.

IBM executives have also warned that many organizations remain unprepared for quantum risks, even as AI-driven attacks increase pressure on existing security systems.

What Should Ordinary Users Do?

If you are not managing an organization's security, you do not need to replace your own encryption manually. Most protection will come through software updates from banks, browsers, messaging apps, operating systems, and cloud providers.

You can still:

• Keep devices, browsers, and apps updated
• Use strong, unique passwords and passkeys where available
• Turn on two-factor authentication
• Avoid reusing old crypto wallet addresses unnecessarily
• Follow updates from your bank, email provider, and wallet provider about post-quantum migration

The most important thing is not to assume “quantum hacking is science fiction.” The timeline is uncertain, but the direction is clear.

---

References

1. Pappalardo, A. et al. Shor’s algorithm compiled on cat qubits: 22 593 qubits for RSA-2048 in 3.14 days. Cryptology ePrint Archive, 2026/2665 (2026). https://eprint.iacr.org/2026/2665
2. Alice & Bob. Alice & Bob Sets Record Low for Quantum Resources Required to Break Internet Encryption. October 29, 2026. https://alice-bob.com/newsroom/alice-bob-sets-record-low-for-quantum-resources-required-to-break-internet-encryption/
3. Alice & Bob. Our Estimate for Factoring RSA-2048 with Cat Qubits: Pushing the Frontier Between Size and Speed. October 29, 2026. https://alice-bob.[STRIPPED 106 bytes]. Gidney, C. and Ekerå, M. How to factor 2048 bit RSA integers in 8 hours using 20 million noisy qubits. Quantum 5, 433 (2021). https://doi.org/10.22331/q-2021-04-15-433
4. Gidney, C. How to factor 2048 bit RSA integers with less than a million noisy qubits. arXiv:2505.15917 (2025). https://arxiv.org/abs/2505.15917
5. Gouzien, É. et al. Performance analysis of a repetition cat code architecture. Physical Review Letters 131, 040602 (2023). https://doi.org/10.1103/PhysRevLett.131.040602
6. Ruiz, D. et al. LDPC-cat codes for low-overhead quantum computing in 2D. arXiv:2401.09541 (2024). https://arxiv.org/abs/2401.09541
7. Google Quantum AI and collaborators. Quantum error correction below the surface code threshold. Nature 638, 920 (2025). https://doi.org/10.1038/s41586-024-08449-y
8. ScienceAlert. Encryption That Protects Your Bank Account Could Fall to a Quantum Computer in 3 Days, Study Suggests. 2026.
9. BankInfoSecurity. Alice & Bob Claims New Benchmark for Hacking RSA in 3 Days. October 30, 2026.
10. G7 / CISA Joint Advisory. Preparing for the Post-Quantum Era. September 3, 2026.
11. Thales. 2026 Thales Data Threat Report: Quantum & AI Trends. 2026.
12. NIST. Post-Quantum Cryptography Standards, FIPS 203, 204, and 205. August 2024.