Skip to the main content.

AFTER THE INVENTORY

 

A Practical Guide To Enterprise Crypto Agility & PQC Remediation

 

A practical playbook for Heads of Cryptography, Security Architecture and Compliance teams: how to turn an inventory of cryptographic assets into a prioritised, funded remediation programme.

 

DOWNLOAD YOUR COPY

EXECUTIVE SUMMARY

 

Cryptography is everywhere in large organisations, and often nowhere when it comes to ownership. Every application that calls an API, terminates TLS, protects a database, signs a release or protects log entries with HMACs depends on keys, certificates, secrets and cryptographic libraries.

Our field experience shows that cryptography can be found across hundreds or thousands of applications. The dependencies accumulate and become hard to track and even harder to control. Each implementation typically followed some reasonable local decision. Together they form an unmanaged layer of enterprise risk, which we at Cryptomathic refer to as “cryptographic sprawl”.

Discovery and inventory establish where cryptography is used, what depends on it and which exposures need attention.

 

 

As individual business flows become sufficiently understood, remediation can begin while discovery continues across the wider organisation.

The work required will vary. Some flows need configuration or software updates; others require changes to key management, application interfaces, infrastructure or external dependencies. Each needs an accountable owner, an agreed treatment and evidence that the change preserves security and business operations.

This paper explains how technical leaders can prioritise that work, establish repeatable approaches to cryptographic change, organise delivery and assess the investment required.

THE PROBLEM HIDING IN PLAIN SIGHT

 

Cryptography does not usually appear on an architecture slide or in the UI, yet it sits underneath almost every business service. Large organisations run application estates measured in hundreds or thousands: one 2025 industry survey reports an average of 897 applications per enterprise [1]. A large share of those applications create, hold or consume cryptographic assets.

897

average applications per enterprise in a 2025 industry survey; nearly half report 1,000 or more

64%

of leaked secrets confirmed valid in 2022 were still exploitable in January 2026, according to vendor research2

47 DAYS

maximum validity of publicly trusted TLS certificates from March 2029, down from 398 days

WHAT A SINGLE APPLICATION CAN DEPEND ON

After the inventory whitepaper tables & images  - Table 1

Table 1: Typical cryptographic dependencies of an enterprise application (illustrative, not exhaustive).

Multiply a handful of such dependencies by a few hundred applications and several environments, and the estate holds tens of thousands of cryptographic objects or assets. Each needs an owner, an approved configuration, a lifecycle and evidence. When every team chooses its own library, key store and rotation practice, the organisation does not have one cryptographic posture, it has many separately managed implementations, with inconsistent oversight and lifecycle practices.

POORLY MANAGED CRYPTOGRAPHY ALREADY HAS A TRACK RECORD

These incidents show how weaknesses in key and certificate management can contribute to serious security failures.

Forgotten keys become master keys

The US Cyber Safety Review Board found that Storm-0558 accessed the mailboxes of senior US government officials with authentication tokens signed by a Microsoft consumer signing key created in 2016. Manual rotation of those keys had stopped entirely in 2021, and no automated alerting on the age of active signing keys was in place3

Expire certificates can switch off defences

In the 2017 Equifax breach, the device monitoring encrypted traffic for the affected application had been inactive for 19 months because its certificate had expired. More than 300 certificates had been allowed to expire, including 79 used to monitor business-critical domains4.

 

Secrets leak faster than they are revoked

Vendor research reports 28.65 million new hard-coded secrets in public GitHub commits during 2025, and finds that 64% of secrets confirmed valid in 2022 were still exploitable four years later

THE PACE OF CHANGE IS ACCELERATING

The CA/Browser Forum has adopted a schedule that cuts the maximum validity of publicly trusted TLS certificates to 200 days from 15 March 2026, 100 days from 15 March 2027 and 47 days from 15 March 2029, with domain-validation reuse falling to 10 days5. The reduction from the previous 398-day maximum to 47 days represents roughly eight times the renewal frequency. Renewal becomes a near-continuous operation. Organisations should assess whether their issuance, deployment and monitoring processes can support that cadence.

PQC then adds new algorithms, larger keys and signatures, new certificate chains and peer compatibility requirements across the same estate. Weak key and certificate lifecycle processes increase migration risk. Addressing those weaknesses during preparation gives teams a more reliable foundation for larger cryptographic changes.

Agentic AI is also accelerating the speed and scale at which cryptographic vulnerabilities can be discovered across an existing IT stack. By analysing code, libraries and configurations, then testing hypotheses, AI agents can expose weaknesses and cryptographic vulnerabilities that have escaped conventional reviews and connect them into exploitable attack paths - making previously unseen exposure actionable.

The UK’s National Cyber Security Centre assesses that AI will enhance vulnerability discovery and exploit development, while further shortening the time between disclosure and exploitation. Cryptography is already within scope: Anthropic reports autonomous discovery and exploitation of flaws in major cryptographic libraries, alongside an agentic workflow that developed and verified an improved attack against the HAWK signature candidate in approximately 60 hours.

 

These results do not break deployed AES or standardised PQC algorithms, but underscore the urgency of discovering, inventorying and remediating hidden cryptographic weaknesses before attackers do.

THREE SYMPTOMS OF UNMANAGED CRYPTOGRAPHY

Invisible. No reliable record of which application uses which key, certificate, library or algorithm, or of who owns it.

Immovable. Algorithm, key location and lifecycle logic are hard-wired into application code, so every policy change becomes a separate project.

Unprovable. No evidence that policy is enforced, that rotation happened or that a retired mechanism is no longer in use.

Crypto agility is the enterprise capability that makes cryptography visible, changeable and provable.

SCOPE OF THIS PAPER

This paper addresses cryptography consumed by enterprise applications and services: business applications, APIs, data platforms, integration layers and software delivery pipelines. It does not address cryptography executed locally on devices, such as operations bound to a trusted execution environment (TEE), trusted platform module (TPM), secure element or microcontroller. Those environments need their own update and attestation design, but their policy can still be defined and evidenced centrally, and many operations that do not need to be device-bound can move to governed central services.

WHY THE WORK MUST BEGIN NOW

 

René Martin puts the issue in operational terms: PQC remediation is a cross-enterprise programme with several essential phases. Management awareness, sponsorship, expertise and funding precede the first operational phase of discovery. The first answer to this challenge is the human factor, not the technology or the algorithms.

The risks derived from cryptography need to be acknowledged by business stakeholders and a change management programme must be established for this to become a priority. Small deployments and early remediation trials then reveal what the organisation must change before it can scale. Building PQC remediation capacity takes time - to address the compliance and risk exposure, immediate action is required.

For financial entities under DORA’s full framework, Regulation (EU) 2024/1774, Article 6(4), already requires provision for cryptanalysis-driven cryptographic updates, or mitigation and monitoring where change is not possible. NIS2’s implementing Regulation (EU) 2024/2690 explicitly requires a cryptographic-agility approach, where appropriate, for the digital and trust-service entities within its scope.

Meanwhile, the EU roadmap recommends starting the transition by end-2026 and migrating high-risk use cases by end-2030. UK NCSC guidance targets discovery and planning by 2028, priority migrations by 2031, and completion by 2035. Crucially, NCSC estimates two to three years for discovery and planning, followed by another two to three years for early migration in large organisations. In certain countries (e.g. Australia, Singapore) the roadmap is even more aggressive.

 

 

Fragmented cryptography and tightly coupled implementations can turn migration into hundreds of separate engineering, interoperability and rollout challenges.

Real crypto agility means being able to change cryptography while preserving security and operations - not simply recording an inventory or supporting additional algorithms. Recommendation to manager: fund remediation teams, policy-controlled abstractions, automated testing and staged migration.

Address current weaknesses during discovery Examining code, configurations and deployed systems will likely uncover unprotected key material, hard-coded credentials and secrets exposed in repositories or logs.

It can also reveal vulnerable cipher suites, obsolete protocols and defective certificate validation that undermine supposedly secure communications. Security critical non compliances are always identified and this also supported in the OWASP Top 10 Application security risks and in the OWASP Cheat Sheet Series.

An inventory helps locate these exposures, but findings must be checked for actual reachability and exploitability - not judged solely by algorithm names or vulnerability counts. Discovery should therefore trigger prioritised remediation immediately, while the broader inventory and post-quantum programme continue.

A CBOM records the exposure; it does not remove it. Quantum readiness must not become an excuse to postpone today’s security fixes.

MAKE COMPLIANCE EXECUTABLE

 

Organisations need evidence that their cryptographic policies are implemented, maintained and reviewed as threats and technology change.

This requires an authorised process for enforcing decisions, managing changes and retaining evidence. The legal provisions below support that argument within their respective scopes. They do not establish one universal PQC mandate or require a particular control platform.After the inventory whitepaper tables & images  - Table 2

Table 2. Selected EU obligations and the agility evidence that supports them.

Applicability depends on entity, sector and national law. Our conclusion is architectural: organisations must be able to control and enforce cryptographic policies. Controlled change and real time visibility is a strong way to implement these expectations.

See how a Tier 1 bank reduced hardware costs and freed up budget for post-quantum readiness.

 

FROM INVENTORY TO THE MOST EXPOSED BUSINESS FLOWS

 

Discovery and inventory tell the organisation which cryptographic assets it holds: keys, certificates, secrets, algorithms, security protocols, cryptographic libraries and the systems that use them. As part of the inventory, the identified cryptographic assets are usually mapped together to draw cryptographic dependencies.

This step is very much important to apprehend the security exposure of a full business flow e.g. signed data set is produced by application A and verified in application C and D. A change requires checking whether applications C and D can accept the new signature profile and updating them where necessary. While looking at this process, it is paramount to assess the business risks if the authenticity or integrity of the data set would be compromised. Discovery shows what you own and where you stand against compliance goals.

It delivers CBOM, maps dependencies, helps define a cryptographic policy. The decisive work is hands-on: finding the business flows that are most exposed and implementing changes that reduce the identified exposure, through governed services and enforceable policy. This helps business leaders set remediation priorities and a recommended way forward.

In practice, making a business flow post quantum ready will include different steps involving architects, service owners, developers, suppliers and crypto specialists:

1) Combine the discovery methods available for the flow: from initiation to archive across applications, gateways, data stores, partners and suppliers. This requires examining network traffic, parsing source code, configurations, protocols etc.

2) Locate every cryptographic decision point: in code, configuration, libraries, HSMs and KMSs, certificates and protocols. Note who can change each one.

3) Build the CBOM: A CBOM (Cryptographic Bill of Materials) is a structured inventory of cryptographic assets - algorithms, keys, certificates and protocols- and their relationships to applications and systems. During discovery, people are often surprised by the amount of cryptographic assets that are difficult to interpret.

4) Before remediation, it is important to assess the exposure of the business flow (not the total assets count)

After the inventory whitepaper tables & images - table 3Table 3: Criteria for selecting the most exposed business flows.

5) Establish what breaks when the algorithm changes: message and certificate sizes, formats, verifiers, stored ciphertext and signatures, partner endpoints and performance.

6) Select the treatment: Decide whether to update, refactor, introduce a shared service, replace, retire or temporarily mitigate the affected implementation. Record the reasons and remaining dependencies.

7) Assign delivery responsibilities: Name the delivery owner, identify required approvals and agree supplier or partner actions and rollout constraints.

8) Implement and test: Check functionality, interoperability, permissions, performance and failure behaviour against agreed acceptance criteria.

9) Deploy and verify: Observe actual cryptographic use, retire obsolete mechanisms where appropriate and document any remaining exceptions.

llustrative example: a payment-initiation API:

  • What the inventory shows: 14 certificates, three RSA keys and two cryptographic libraries attached to the service.

  • What the flow analysis reveals: payment instructions are signed with an RSA key stored in application configuration; the signatures are archived for ten years for dispute handling; a partner bank verifies them with its own library; mTLS terminates at a shared API gateway used by 40 other services.

  • The exposure of the flow: This is a high-priority flow combining potential signing-key exposure, ten-year evidentiary risk, partner-dependent migration and a shared gateway whose disruption could affect 41 services.

A CRYPTO-AGILE ARCHITECTURE TO FACILIATE REMEDIATION

 

When algorithm choices, key locations and lifecycle rules are tightly coupled to application code, cryptographic changes require coordinated work across application teams. A shared cryptographic service can centralise selected responsibilities and provide consistent interfaces for approved operations. Its value depends on which responsibilities it assumes and how the remaining dependencies are managed.

After the inventory whitepaper tables & images  - Image 12-1

Figure 1. Crypto agility reference architecture: applications send cryptographic commands to a central service that verifies each request against policy, uses governed keys and executes across HSMs, while service administration registers applications and assigns policy and keys.

The architecture in Figure 1 has different responsibilities. Business apps express operations through approved interfaces. A Service administration defines application access control, manages policies (including algo change), manages key and certificate association (including rekey, rotation or reshape). It also typically monitors execution and system health through integration with central SIEM or ITSM. The CrystalKey 360 Services ensure smooth execution under high performance requirements.

The application no longer maintains cryptographic assets locally but CrystalKey 360 Services maintain them centrally and can produce a live CBOM. This reduces drastically the cryptographic exposure of the application as it simply knows the cryptographic intent and not the cryptographic primitives.

This is made possible using CQL is CrystalKey 360’s cryptographic intent interface, enabling applications to express what they need to achieve - such as encrypting data or signing a document - while centralising control over how cryptography is performed. Whereas standard interfaces such as PKCS#11, JCE and CNG typically require applications to select cryptographic mechanisms and parameters, CQL separates application intent from the management of keys, algorithms and cryptographic primitives. This decoupling enables effective crypto agility: cryptographic choices can evolve centrally, reducing the need for application changes. It also provides central visibility into cryptographic usage and consistent policy enforcement across connected applications, helping organisations govern their cryptography and adapt to new threats and requirements.

After the inventory whitepaper tables & images 2

Let us take an example: e.g. an invoice generation application that

a) keeps customer data encrypted at rest,

b) generates e-invoices in the EN 16931 semantic standard,

c) seals them twice using a XAdES enveloped seal on the UBL XML invoice and a PAdES seal on a human-readable PDF,

d) uploads the sealed files into a target folder over an HTTPS channel with mutual TLS authentication.

If we consider the Crypto Agile Architecture above, every operation that needs a secret or a private key runs inside CrystalKey 360 (CK360) through CQL (Crypto Query Language) on the Crypto RESTAPI.The program never stores, generates orreceives a private key, a secret key or a data-encryption key.

After the inventory whitepaper tables & images  - table 4

Moving to post-quantum cryptography is then tremendously easier with the crypto agile architecture as it only required a central policy update and key reshape and the new primitives will be inherited to protect the assets in our example:

After the inventory whitepaper tables & images

Key material stored locally: none. The CBOM references CK360 keys by name only. The workspace holds ciphertext, public certificates and sealed invoices. The demo and the tests check that no file contains a private key.

AGENTIC AI: POST-QUANTUM REMEDIATION AT THE SPEED OF THE ESTATE

 

The hundred or thousand application problem is ultimately an engineering challenge. Moving direct cryptographic calls behind a governed interface means editing code in hundreds of repositories, each with its own language, framework and owner. AI coding agents change the economics of that work. They must not, however, become a new source of unmanaged cryptography.

After the inventory whitepaper tables & images - image 3

Figure 3. Agentic AI can help remediate existing applications and generates new ones, leveraging CK360 MCP enabled integration.

Existing applications need conversion. A code scanner or discovery product detect the cryptographic assets, builds compliance mapping and exposes weaknesses, vulnerabilities or non compliant items. The team prioritises exposed flows; the agent then rewrites direct cryptographic calls into intent-based CQL calls and requests the required consumer identity, credentials, keys and policy binding.

New applications need prevention. The agent generates CQL-first code from the first commit, with no provider-bound APIs, hard-coded algorithms or application-owned key lifecycle and can extract the CBOM. Using approved integration patterns from the outset can reduce new unmanaged cryptographic dependencies.

KEEP THE AGENT INSIDE THE POLICY

A Model Context Protocol (MCP) server forms the controlled action layer. It exposes only approved operations, under scoped identity, with admissible parameters, human approval where required and an audit trail. It never gives the agent raw HSM, key-store or provider authority. Cryptomathic has demonstrated this pattern with coding agents, an MCP server and CK360 APIs.

After the inventory whitepaper tables & images - table 7

Illustrative CQL request and policy parameters.

The concept has two properties.

Least Crypto Privilege: an application receives only the cryptographic authority its business function needs. It does not choose algorithms, hold unmanaged credentials or decide when keys rotate.

Intrinsic Crypto Agility: a key rotation is straight-forward. An algorithm change, e.g. RSA to ML-DSA, becomes a policy and lifecycle event rather than another round of code changes across the estate.

AI also helps beyond code. It can correlate discovery evidence, draft flow maps, propose test cases and generate runbooks from approved patterns. Human review, deterministic tests and secure handling of code and secrets remain essential.

USE THE TRANSITION TO LOWER RECURRING COST

 

When each application builds its own cryptographic integration, the enterprise pays repeatedly for infrastructure, specialist operations, onboarding and every later change. Shared, policy-governed services remove that duplication.

Cryptomathic estimates savings in three classes.

1) Hard cash is removable HSM, KMS, tooling, hosting and external-service spend.

2) Conditional staffing savings reflect reduced central crypto-operations effort, counted as cash only with an approved cost-removal or hiring-avoidance plan.

3) Application-team capacity value is time released in application and project teams, of which only a stated share is recognised.

A WORKED CASE: A UNIVERSAL BANK WITH 600 APPLICATION

The model assumes a universal bank with 600 crypto-consuming applications and 12 central crypto-operations FTEs at EUR 150,000 loaded cost. Benefits start in month 9. The figures are illustrative and you are welcome to come to us for a detailed analysis.

EUR 2.24M

annual run-rate cash savings, of which EUR 1.26M is provisional central-team labour

171%

five-year economic ROI: EUR 12.38M benefits less EUR 4.75M cost leaves EUR 7.81M

12 MONTHS

economic payback; on cash savings alone, about 14 months and 112% ROI

After the inventory whitepaper tables & images  - TABLE 8

Table 6. Annual run-rate benefit in the ROI calculator’s reference case.

Commission enough governance to support safe decisions, then demonstrate one end-to-end remediation and a reusable pattern. Broader discovery continues throughout.

A 6-MONTH PLAN TO PROVE THE CHANGE PATH

MONTHS 1-3: SELECT THE MOST EXPOSED FLOWS

Appoint an executive sponsors and Cryptographic Centre of Excellence (CCoE) lead. Approve provisional policy and exception authority. Apply the exposure criteria to select three to five flows, covering confidentiality and authenticity. Run targeted discover on those flows and establish the cost baseline.

HERO02

MONTHS 2-5: DESIGN THE REMEDIATION

Work each flow end-to-end with its owners. Select the treatment, approve the first policy profiles and build the first reusable pattern with interoperability, performance, negative and downgrade tests. Pilot agent-assisted remediation.

HERO02

MONTHS 3-6: DEMONSTRATE CAPABILITY

Run the remediated flow in an authorised environment and release progressively when acceptance criteria pass. Observe actual cryptographic use. Capture closure evidence, compare measured effort, including AI throughput, with the budget and approve the next wave from those results.

HERO02

BUILD CHANGE AND TRY REMEDIATION EARLY IN THE PROCESS

Shared crypto services bring compliance and make the transition affordable by bringing efficiency (less onboarding time, faster to develop, more efficient use of scarce cryptographic talent and also fewer HSMs and KM solutions). Consolidating fragmented cryptography removes duplicated engineering effort, reduces infrastructure cost and audit effort, and verified savings can fund later waves. It is also crucial for ensuring real time overview and control over cryptographic assets and it also brings the ability to act swiftly upon indication of compromise, which is precisely what regulators expect.

THE DECISION THAT CHANGES THE PROGRAMME

 

Make crypto agility a funded enterprise capability. Use discovery to see the estate, then put the effort where exposure is highest: hand-on remediation of the business flows that matter most, delivered through governed services and accelerated by AI under policy. 

PQC then becomes a sequence of governed changes with measurable outcomes. The lasting return is an organisation that can respond to the next cryptographic weakness with less delay, less repeated engineering and evidence that its protection remains effective.

cryptomathic_symbol_core_negative_transparent
AN ENTERPRISE ARCHITECTURE GUIDE pOST-QUANTUM READINESS

An Enterprise Architecture Guide For Post Quantum Readiness

This guide explains how organisations can prepare for post-quantum cryptography by building the governance and operating model needed to manage cryptographic change.

Read the guide

 

CRYPTO-AGILITY FROM CRYPTOGRAPHIC INFRASTRUCTURE TO ENTERPRISE CONTROLthumbnail

Crypto-Agility: From Cryptographic Infrastructure to Enterprise Control

This whitepaper explains the operational challenges of cryptographic visibility, ownership and more, while outlining the capabilities needed to support security demands and future PQC migration.

Read the whitepaper

 

QUANTUM SECURE KEY MANAGEMENT Thumbnail-3

Quantum Secure Key Management: PQShield and Cryptomathic's CrystalKey 360

This guide provides a combined solution to quantum challenges through the strategic technology alliance between PQShield and Cryptomathic, diving into Cryptomathic's CrystalKey 360.

Read the brief

 

A PRACTICAL NEXT CONVERSATION 

Talk to a Cryptomathic expert about identifying your most exposed flows and proving your first remediation.

TALK TO AN EXPERT TODAY