Pick a service your team shipped this year. Which algorithm signs its tokens? What key size? Which line of code made that choice, and who wrote it?
Most teams can't answer without a week of digging. For years that was fine. Cryptography worked, nobody touched it, and the libraries did the hard part.
That is about to change. Quantum computers are expected to break RSA and elliptic-curve cryptography, the public-key math behind TLS, code signing, single sign-on, and most of the internet. Governments have now set dates for moving off them. And every migration plan, in every guideline, starts with the same first step: find out what you have.
That inventory has a name: a Cryptographic Bill of Materials, or CBOM.
How does quantum computing threaten current encryption like RSA and TLS?
In 2019, researchers estimated that breaking RSA-2048 would take a quantum computer with about 20 million physical qubits, running for about eight hours. In May 2025, Google's Craig Gidney cut that estimate to fewer than one million qubits, running for less than a week. In October 2026, Alice & Bob published an estimate of 22,593 physical qubits and about three days on its cat-qubit architecture, which assumes qubits that almost never suffer bit-flip errors.
Those numbers fell mainly because of better algorithms and better error correction, and in Alice &Bob’s case a new type of qubit. These papers are estimates, not demonstrations. No quantum computer can do this today, and serious engineering problems remain. But the direction is hard to ignore. In the Global Risk Institute's Quantum Threat Timeline Report 2025, published in March 2026, 26 experts put the chance of a cryptographically relevant quantum computer within 10 years at 28% to 49%.
You don't need to pick a date for "Q-Day" to see the problem. You only need to compare two timelines: how long until the threat is real, and how long your migration will take.
What is a "harvest now, decrypt later" attack?
The threat doesn't start when the first large quantum computer comes online. It started already.
In a "harvest now, decrypt later" attack, an adversary records encrypted traffic today and stores it until it can be decrypted. NIST calls this one of the main reasons the transition is urgent. If your systems handle data that must stay confidential into the 2030s (health records, financial data, intellectual property, government information), then any of it sent today over quantum-vulnerable key exchange is already exposed.
For that data, the deadline isn't 2030. It's the day it was sent.
What are the NIST post-quantum cryptography deadlines and requirements?
Governments started putting dates on post-quantum migration in 2022. Over the past two years, most major economies have set their own deadlines, and the US has pulled its most sensitive systems forward to 2030.
- NIST published its final transition report, IR 8547, in October 2026. It deprecates quantum-vulnerable algorithms such as RSA and ECDSA after 2030, at every key size, and disallows them under NIST standards after 2035.
- The replacements are already standardized: ML-KEM (FIPS 203), ML-DSA (FIPS 204), and SLH-DSA (FIPS 205).
- US federal agencies have had to submit a yearly inventory of the quantum-vulnerable cryptography in their most sensitive systems since May 2023, under OMB memo M-23-02. In June 2026, OMB memo M-26-15 went further: it asks agencies to keep a continuously updated inventory of all cryptographic assets that feeds a central cryptographic bill of materials. Executive Order 14412 now requires federal high-value assets and high-impact systems to use post-quantum key establishment by December 31, 2030 and post-quantum digital signatures by December 31, 2031.
- US national security systems follow the NSA's CNSA 2.0, which requires new acquisitions to be compliant from January 1, 2027, unless otherwise noted.
- The European Union's coordinated roadmap, published in June 2025, asks member states to complete first steps, including building cryptographic inventories, by the end of 2026, migrate high-risk use cases by the end of 2030, and migrate medium-risk use cases by the end of 2035, with low-risk systems moved as far as feasible.
Similar timelines exist in the UK, Canada, Japan, and many other countries, and Australia's is earlier still: the end of 2030. The details differ. The first step doesn't. NIST's own announcement points beyond an organization's own systems to its "entire cryptographic supply chain."
Regulators are also starting to name the CBOM itself. Executive Order 14412 directs CISA and NIST to publish minimum elements for a cryptographic bill of materials. OMB's M-26-15 asks agencies to feed their inventories into a central CBOM. The EU roadmap recommends a standard format such as CBOM, and Australia's signals directorate recommends CBOMs too.
Why finding your cryptography is harder than it sounds
If cryptography lived in one place, this would be easy. It doesn't.
It's spread across thousands of files written over many years. It's in your own code, in the libraries you depend on, in configuration files, and in certificates loaded at runtime. The person who picked the cipher may have left years ago. And the code is changing faster than ever, now that AI coding agents write a growing share of it. Every new line can add a new place where cryptography is used.
Searching the repository for "RSA" won't find all of it either. Look at this Java snippet:
String transform = config.get("token.cipher"); // "RSA/ECB/PKCS1Padding"
Cipher cipher = Cipher.getInstance(transform);
cipher.init(Cipher.ENCRYPT_MODE, publicKey);The algorithm name never appears next to the call that uses it. It's also weak in a way that has nothing to do with quantum computers: RSA with PKCS#1 v1.5 padding is a known risk today. A text search misses it. A developer survey misses it. A spreadsheet built last quarter is already out of date.
This is why most teams can't answer the questions at the top of this post. It's not a skill problem. It's a visibility problem.
What is a Cryptographic Bill of Materials (CBOM)?
A Cryptographic Bill of Materials (CBOM) is a machine-readable inventory of the cryptography a piece of software uses, and where it uses it.
The main open standard for CBOMs is CycloneDX, the open source bill of materials standard from OWASP that is also published as Ecma International standard ECMA-424. CBOM support arrived in CycloneDX 1.6, and version 1.7 expanded it with a shared registry of algorithm families and elliptic curves, so different tools describe the same cryptography the same way. The OWASP Authoritative Guide to CBOM covers the full data model.
A CycloneDX CBOM records four kinds of cryptographic assets:
- Algorithms, such as AES-256-GCM, RSA-2048, SHA-1, or ML-KEM-768, with properties like mode, key size, and security level.
- Protocols, such as TLS versions and the cipher suites they allow.
- Certificates, such as the X.509 certificates that prove identity and carry public keys.
- Related cryptographic material, such as keys, initialization vectors, nonces, and salts. A CBOM describes what kind of material is used, not the secret values.
Each entry can also carry evidence: the file and line where the asset was found. Here is a simplified example for the snippet above:
{
"type": "cryptographic-asset",
"name": "RSA-2048",
"cryptoProperties": {
"assetType": "algorithm",
"algorithmProperties": {
"algorithmFamily": "RSA-PKCS1",
"primitive": "pke",
"parameterSetIdentifier": "2048",
"cryptoFunctions": ["encrypt"],
"classicalSecurityLevel": 112,
"nistQuantumSecurityLevel": 0
}
},
"evidence": {
"occurrences": [
{ "location": "src/main/java/com/acme/payments/TokenEncryptor.java", "line": 42 }
]
}
} The last property tells the story. A NIST quantum security level of 0 means the algorithm meets none of NIST's post-quantum security categories: a large enough quantum computer could break it. And the evidence tells you exactly where to go to fix it.
What is the difference between a CBOM and an SBOM?
If you already produce a software bill of materials (SBOM), don’t assume it covers your cryptography. The two answer different questions.
An SBOM might tell you that an application ships Bouncy Castle. Only a CBOM tells you that it uses RSA-2048 for encryption on line 42 of the payments service. Both use CycloneDX, so they fit side by side.
What a CBOM lets you do
An inventory isn't the goal. It's what makes every next step possible.
- Find what's at risk. Separate quantum-vulnerable algorithms like RSA and elliptic curves from those that are already safe, and from those that are broken today, like MD5 and SHA-1.
- Decide what to fix first. Rank work by how long the protected data must stay secret, how sensitive it is, and which regulations apply.
- Hand developers a real task. "Replace RSA somewhere in this service" is a project. "Change the call on line 42" is a ticket.
- Show progress with evidence. A standard file that updates as code changes is something auditors and regulators can read. A spreadsheet isn't.
- Build crypto-agility. Once you know where cryptography lives, swapping algorithms becomes a planned change instead of an emergency.
One caution: a CBOM doesn't make you compliant on its own. It's evidence and a planning input. Compliance comes from the migration work it guides.
What a good CBOM needs
Not every inventory is equally useful. When you evaluate how to produce one, look for five things.
- Built from code, not surveys. The inventory should come from analyzing what the code actually does, including values that flow in from configuration.
- Consistent and repeatable. The same code should produce the same inventory every time, so changes in the CBOM reflect changes in the code.
- Explainable down to the line. Every entry should point to the file and line where it was found. If a developer can't find it, they can't fix it.
- Current with every change. Generate it in the pipeline, alongside the analysis that already runs on every branch, so it never goes stale between audits.
- In an open standard. Use CycloneDX so the CBOM works with your SBOM, your governance tools, and whatever comes next.
Where Sonar stands
Every line of code should be verified before it ships, whether a developer or an AI agent wrote it. Cryptography is part of that code.
SonarQube already flags weak cryptography in code, including weak hashing, insecure cipher modes, and padding, short keys, and outdated TLS versions. With SonarQube Advanced Security, teams can also export an SBOM in CycloneDX or SPDX format to track the open source components they depend on.
The next step is knowing exactly which cryptography that code uses, and where. That’s why we’re bringing CBOM generation to SonarQube, built from the same analysis that already runs on every branch. It will include cryptographic posture management: one view of the cryptography across all your projects, checked against a crypto baseline and post-quantum readiness policies, so you can see the gaps and track the migration.
CBOM generation and cryptographic posture management are now in early access on SonarQube Cloud with a select group of customers. If you’d like to join, contact us for a demo and we’ll show you what it finds in your code.
The first federal deadlines are about four years away, and replacing cryptography across a large codebase takes years. Start the inventory now, while you still have time to plan the work instead of rushing it.

