![]()
Most security teams only think about certificates when one expires and a customer-facing service goes dark. That reactive habit is becoming expensive. Public TLS certificate lifetimes are shrinking every year, device fleets keep growing, and auditors now ask who owns each key. This is where a planned enterprise PKI programme, the kind Schutz IT designs and runs for organisations in Australia, the United States and Europe, changes the conversation from firefighting to routine operations. This guide explains what has changed, where most teams struggle, and how to build a certificate practice that keeps working as volumes rise.
Why Certificate Management Became a Board-Level Risk
In April 2025 the CA/Browser Forum approved a phased reduction in the maximum validity of public TLS certificates. Since 15 March 2026 the ceiling has been 200 days. It drops to 100 days in March 2027 and to just 47 days in March 2029, with domain validation reuse periods shrinking alongside it.
On paper that is a small policy change. In practice it multiplies the number of renewals a security team must perform. An organisation that renewed 400 public certificates once a year will be renewing them roughly every six to seven weeks by the end of the decade. Spreadsheets, calendar reminders and a single administrator who "knows where the certificates live" cannot keep up with that cadence.
Internal certificates add a second layer of load. Laptops, mobile devices, Wi-Fi access, VPN gateways, code signing, email encryption and IoT sensors all depend on certificates issued by a private Certificate Authority. Each one has its own expiry date, its own key, and its own owner who may have left the company two years ago.
The outcome is predictable: outages caused by an expired certificate, failed audits because nobody can produce a full inventory, and weak keys that stay in production long after they should have been retired.
The Building Blocks of a Modern PKI Programme
A well-run PKI is less about buying a product and more about putting four disciplines in place.
1. A complete certificate inventory. You cannot manage what you cannot see. Discovery scans across networks, cloud accounts, load balancers and endpoint management tools should produce one register that shows every certificate, its issuer, its expiry, its key length and its business owner.
2. Clear ownership and policy. Every certificate needs a named owner and a policy that defines permitted algorithms, key sizes, validity periods and approval steps. Policy is what allows automation to run safely without a human signing off on each request.
3. Automated enrolment and renewal. Protocols such as ACME for web servers, EST and SCEP for devices, and native integrations with Microsoft Intune and JAMF allow certificates to be issued and renewed without manual effort. Automation is the only realistic answer to 47-day lifetimes.
4. Protected keys and a trusted hierarchy. Root and issuing CA keys belong in hardware security modules, with a documented hierarchy, revocation process and disaster recovery plan. A compromised issuing CA can undermine every identity it has ever signed.
Teams usually choose between three operating models. A managed PKI hands day-to-day operation to a specialist partner. A private PKI gives the organisation a dedicated Certificate Authority under its own governance. A cloud PKI, often described as PKI as a Service, delivers the same capability with automated enrolment and hardware-backed key protection, without on-premises servers to patch. The right choice depends on how much control, ownership and operational responsibility the business wants to keep in-house.
Certificate Lifecycle Management: Where Most Teams Get Stuck
Even organisations with a mature CA often stall at lifecycle management. The common sticking points are consistent across industries.
- Shadow certificates. Development teams request certificates from public CAs with a company card, and those certificates never reach the central inventory.
- Manual renewals on critical systems. Legacy appliances and mainframe connections often cannot use ACME, so renewals remain manual and fragile.
- Revocation that nobody tests. Certificate revocation lists and OCSP responders are configured once and then forgotten until an incident proves they do not work.
- No link to identity systems. Device certificates are issued, but nobody checks whether the user or device is still active in the directory.
A practical starting point is a 90-day plan. In the first month, run discovery and build the inventory. In the second month, assign owners, define policy and fix the certificates closest to expiry. In the third month, automate renewal for the highest-volume categories, usually web servers and managed endpoints, and set up alerting for everything that cannot yet be automated. By the end of the quarter, renewals should be a scheduled process rather than an emergency.
Machine Identity and Customer Identity Belong in the Same Plan
PKI secures machine and code identity: servers, devices, applications and signed software. Customer Identity and Access Management (CIAM) secures human identity: the customers and partners who sign up, sign in and manage their accounts in your applications. They are different disciplines, but they meet at several points, and planning them separately creates gaps.
Passwordless sign-in is the clearest example. Passkeys and FIDO2 security keys rely on public key cryptography, the same foundation as PKI. Certificate-based authentication for partner portals, mutual TLS between APIs, and token signing keys for OpenID Connect all depend on keys that must be generated, stored, rotated and retired correctly. If your certificate practice is weak, your customer identity platform inherits that weakness.
A successful CIAM implementation project therefore starts with strategy rather than configuration. That means defining the identity data model, data residency and compliance requirements, the migration approach for existing users, and which sign-up, social login and progressive profiling journeys need custom logic. Platforms such as Microsoft Azure AD B2C and Entra External ID, Okta Customer Identity and Auth0 each handle these needs differently, and the right choice depends on scale, standards support and how well the platform fits the existing technology stack.
When PKI and CIAM are planned together, the organisation gains one view of every identity it trusts, human or machine, and one set of policies for how those identities are created, verified and removed.
A Practical Checklist for the Next Six Months
Use the following checklist to judge where your organisation stands today:
- Can you produce a full certificate inventory, with owners, in under a day?
- Are at least 80 percent of public TLS certificates renewed automatically?
- Are root and issuing CA keys stored in hardware security modules with tested recovery?
- Is there a written policy for algorithms, key sizes and validity periods?
- Do device certificates get revoked automatically when a device or user leaves the directory?
- Has your customer identity platform been reviewed against your passwordless and MFA roadmap?
- Is there a named owner for the cryptographic keys that sign your identity tokens?
If the answer to three or more of these questions is "no" or "not sure", the shrinking certificate lifetimes coming in 2027 and 2029 will expose those gaps quickly. The good news is that each item can be fixed with a clear plan, the right automation and sensible governance, and none of them requires replacing systems that already work.
Final Thoughts
Certificates and customer identities are now operational infrastructure, not one-off projects. Organisations that invest in inventory, policy and automation today will handle 47-day certificates and passwordless customer journeys as routine work, while those that wait will keep facing outages and audit findings. If you would like an assessment of your current PKI or customer identity setup, you can contact the Schutz IT team to discuss your environment and priorities.