Quantum-Safe Security: Why It Will Become an Obligation, Not a Hype

Every time a new security topic appears, noise comes with it. Quantum computing has suffered from this for years: distant promise, vendor slides, fear sold as a service. I understand the distrust of those who lead technology, because we have already seen this film with several acronyms.

The difference is now concrete. In August 2024, NIST finalized the first post-quantum cryptography standards. It left the field of speculation and entered the field of published standards, with associated government deadlines. When this happens, it stops being hype and becomes planning.

This text is the starting point of the series. I want to make clear what it means to be quantum safe, why the clock is already ticking, and what a technical leader needs to decide before the decision is made by an auditor or a customer.

What does “quantum-safe” mean?

Quantum-safe (or post-quantum) describes systems whose cryptography resists attacks from a cryptography-relevant quantum computer. It's not a unique technology, it's a property that you want in your systems.

The problem is specific. Much of the encryption that protects the internet today relies on two public key algorithms: RSA and elliptic curve cryptography (ECC). They support browser locking, key exchange in secure connections, software signatures, certificates, authentication. Their security comes from mathematical problems that a classical computer takes an impractical time to solve.

A sufficiently capable quantum computer changes this account. Known quantum algorithms solve exactly the problems that underlie RSA and ECC. When this machine exists on a useful scale, the public key that protects your data is no longer protected.

Notice what this doesn't achieve. Symmetric encryption, such as AES, and hash functions do not fall into the same shape. They suffer a weakening that is compensated by increasing the size of the key. The sticking point is public-key cryptography, and that's what the new standards focus on.

Why the clock is already ticking

The natural objection is: if the quantum computer doesn't yet exist on this scale, why bother today? Because the threat doesn't wait for the machine to be ready.

There is an attack called "harvest now, decrypt later", harvest now and decrypt later. The attacker doesn't need to break anything today. It captures and stores encrypted traffic and data now, knowing it will be able to decrypt when it has quantum capabilities. Everything you transmit with vulnerable protection is a deposit waiting to be withdrawn.

This reverses the risk logic. The question stops being “when will the quantum computer arrive” and becomes “how long does my data need to remain secret”. Records, contracts, intellectual property, citizen data, state secrets: if the validity of confidentiality exceeds the horizon of arrival of quantum computing, the data is already exposed. This mechanism deserves its own article, and I will deal with it in detail later in the series.

What NIST delivered in 2024

The finalization of standards by NIST is the milestone that transforms the subject into a project. Three algorithms came out as federal standards.

ML-KEM, published as FIPS 203, takes care of key encapsulation, the mechanism that establishes a shared secret key over a connection. It was known during the competition as CRYSTALS-Kyber. It is the direct replacement for key exchange which today relies on RSA and ECC.

ML-DSA, published as FIPS 204, is the digital signature standard, used to ensure authenticity and integrity: certificates, software updates, signed documents. CRYSTALS-Dilithium vine.

SLH-DSA, published as FIPS 205, is a more conservative hash-based signature, designed as a backup alternative because it relies on different mathematical assumptions. Diversity of fundamentals is a defense in itself.

Having published standards matters because it unlocks everything that comes after: libraries, certifications, contractual requirements, audits. The supplier is now able to commit to something concrete, and the customer is now able to charge.

The deadlines that make this inevitable

Standard without deadline becomes intention. And deadlines exist.

The NSA, per the CNSA 2.0 guideline, targets 2030 for national security systems to adopt post-quantum algorithms. NIST signals the deprecation of vulnerable algorithms by 2035. These are American government references, but the severity of the dates drives the market: suppliers, certifications and regulatory bodies tend to align with them.

For those who lead technology in Brazil, these dates are not foreign folklore. Libraries, operating systems, cloud providers, and hardware manufacturers will move around this calendar. When a browser starts to require post-quantum algorithms, or a provider deprecates a set of ciphers, the effect reaches your infrastructure without asking for permission.

Add to this the LGPD. The obligation to protect personal data by appropriate technical measures does not freeze in time. If there is a known method to compromise the protection of long-term data, maintaining vulnerable algorithms without a plan becomes a weakness that is difficult to justify in the face of an incident.

What changes in practice for those who lead

Migrating to post-quantum cryptography is not changing a library on a Friday night. It is an inventory and architectural exercise.

First, you need to know where you use encryption. Most organizations don't know. Keys, certificates, libraries, protocols and dependencies are spread across applications, integrations and vendors, often without a central registry. You don't migrate what you can't see, and this cryptographic inventory is the prerequisite for everything.

Second, you want crypto-agility: the ability to change the cryptographic algorithm without rewriting the entire system. Whoever tied figures into the code will suffer. Those who have isolated cryptography behind well-defined layers trade with much less pain. This is projected, not improvised.

Third, there is implementation care. Post-quantum algorithms are new to operational practice, have different key and signature sizes, and rushed implementations introduce their own flaws. The sensible approach combines the classical algorithm with the post-quantum one in hybrid mode during the transition, maintaining the current protection while adding the new one.

None of this requires panic. It requires starting with diagnosis: finding out which data has a long shelf life, which systems depend on RSA and ECC, and which vendors already have a plan. The rest of the series delves deeper into each of these fronts.

If you lead technology or security, start crypto inventory before the end of this quarter. It is not a sprint project, and whoever maps it first decides calmly instead of reacting under audit pressure.