Preparing Windows for the Quantum Age: Microsoft Hardens Windows 11 Preview with New Encryption

[AleksandarK]

To defend regular users from bad actors wielding quantum computing power like Majorana 1, Windows 11 Insider Preview now includes built-in support for post-quantum cryptography (PQC), giving developers and security teams early access to algorithms designed to withstand the capabilities of future quantum computers. Available in Canary Channel Build 27852 and above, this update integrates two new schemes, ML-KEM for key exchange and ML-DSA for digital signatures, directly into the Cryptography API: Next Generation (CNG) and certificate management functions. ML-KEM addresses the "harvest now, decrypt later" threat model, in which adversaries collect encrypted data today to decrypt it once quantum hardware has advanced. Microsoft offers three levels of ML-KEM security: a Level 1 option that produces 800-byte ciphertexts and a 32-byte shared secret; a Level 3 configuration with 1,184-byte ciphertexts and the same 32-byte secret; and a Level 5 tier that increases ciphertext size to 1,568 bytes while keeping the shared secret at 32 bytes. These parameter sets allow organizations to balance performance and protection according to their threat models and operational requirements.




ML-DSA complements key exchange by providing quantum-resistant digital signatures. Three strength tiers are available for trial: Level 2 yields compact public and private keys of 1,312 and 2,560 bytes, respectively, with 2,420-byte signatures; Level 3 increases key sizes to 1,952 and 4,032 bytes with 3,309-byte signatures; and Level 5 offers the highest assurance with 2,592-byte public keys, 4,896-byte private keys, and 4,627-byte signatures. While these larger footprints may impact storage and transmission, they deliver stronger guarantees against future signature-forging attacks. Microsoft recommends running these post-quantum algorithms in hybrid mode alongside classical counterparts such as ECDH, RSA, or ECDSA to ensure in-depth defense.

Early adopters can install, import, and validate PQC-based certificates within the Windows certificate store and measure the impact on handshake latency, certificate size, and API integration. On the Linux side, SymCrypt-OpenSSL 1.9.0 introduces equivalent hybrid key-exchange capabilities for TLS based on the latest IETF drafts. This unified preview across Windows and Linux provides a consistent experimentation, feedback, and performance-tuning environment. Microsoft plans to extend PQC support to the native Windows TLS stack (Schannel), Active Directory Certificate Services, and Intune's certificate connector.

View at TechPowerUp Main Site | Source

 

[Chaitanya]

Does this also includes backdoors for spywares of various governement agencies and M$ themselves?

 

[NoneRain]

Does this also includes backdoors for spywares of various governement agencies and M$ themselves?

-----------------
Thanks MS, now we're protected against tech we are far from achieving; that's exactly what Win needed right now.

 

[mb194dc]

Completely pointless but ads to the quantum hype machine. That's all the tech industry is now. Reams of bullshit.

Meanwhile msft and many others like crowdstrike don't even have QA teams. So we get frequent borkages.

 

[theouto]

Holy fuck, how paranoid can one company get? We won't get there in some decades, Microsoft, it's alright, don't worry about it.

 

[zmeul]

the only thing MS protects is users from their own data

the mandatory BitLocker activation on new installations is one of the biggest "fuck you" to home users

 

[Timbaloo]

Windows 11 is already quantum ready. Will explorer respond or will it not respond?

 

[FierceRed]

Does this also includes backdoors for spywares of various governement agencies and M$ themselves?

Yes.

Windows 11 is already quantum ready. Will explorer respond or will it not respond?

No, it will not respond. It already responded in all the other timelines so your explorer owes you nothing.

 

[R-T-B]

Holy fuck, how paranoid can one company get? We won't get there in some decades, Microsoft, it's alright, don't worry about it.

When you are harvesting data for future decryption "some decades" becomes relevant. Plus I doubt we are that far off, frankly.

But what do I know. I'm probably just the only security researcher on this forum.

 

[Onasi]

@R-T-B
Actual non-meme question - what ARE yours and overall industries prognosis on how long it would take for Quantum-based decryption algorithms to become an actually valid threat vector? Something we will see in a decade? Less? What are the main difficulties in hardening an end-user system (be it enterprise or consumer) for that?

 

[R-T-B]

@R-T-B
Actual non-meme question - what ARE yours and overall industries prognosis on how long it would take for Quantum-based decryption algorithms to become an actually valid threat vector? Something we will see in a decade? Less? What are the main difficulties in hardening an end-user system (be it enterprise or consumer) for that?

I really have no real measuring stick to go by, but "based on feels" I'd say we have about 5-10 years before nation states have the power to break commonly used encryption.

It will be much longer for individuals of course.

As for the hardening? You really just switch algorithms. Things like AES break easily but there are other encryptions that are far more resiliant by nature. The real issue is the reliance on AES and especially, SSLs use of it.