Explore our new Cybersecurity Terms Glossary, explained by our experts.

View the Glossary View the Glossary
PRIVIA

Cybersecurity Terms Glossary

What Is a Self-Signed Certificate?

A self-signed certificate is signed by its own private key rather than by a certificate authority, so it carries encryption but no independent proof of who is on the other end.

Illustration for the cybersecurity term Self-Signed Certificate
Home / Blog / Cybersecurity Terms / What Is a Self-Signed Certificate?
17 November 2026

What Is a Self-Signed Certificate?

What Is a Self-Signed Certificate?

A self-signed certificate is an X.509 certificate in which the issuer and the subject are the same entity: the certificate is signed by the very private key it advertises. It still carries a public key, a subject name, and validity dates, and a TLS connection using one is encrypted exactly as any other. What it does not carry is an independent assertion that the holder is who the name claims.

That single missing property explains every browser warning, every failed API call, and every awkward conversation about certificates on an internal network.

How It Differs From a CA-Issued Certificate

When a public certificate authority issues a certificate, it first validates control of the domain, then signs with a key whose root is already present in operating system and browser trust stores. A client can therefore build a chain of trust from the server certificate up to a root it already trusts. A self-signed certificate has a chain of length one that ends nowhere, so the client has nothing to compare against except a fingerprint someone gave it out of band.

Three practical consequences follow. There is no revocation mechanism, because no CA publishes a CRL or OCSP response for it. There is no Certificate Transparency record, so nobody outside your organisation can detect its issuance or misuse. And there is no issuance policy, so a developer can generate one in ten seconds with any key size, any name, and a ten-year validity period.

Where They Are Perfectly Legitimate

The Risks Nobody Budgets For

The damage is rarely the certificate itself. It is the habits built around it:

Better Options

For anything with a public name, use a publicly trusted certificate. They are free from Let’s Encrypt and fully automatable through ACME. For internal services, stand up a private CA—Microsoft AD CS, step-ca, or a managed private CA—and distribute its root through Group Policy or your MDM. A private CA gives you a single trust anchor, central revocation, name constraints that stop it issuing for domains you do not own, and the same automation as the public ecosystem. For workload identity inside a cluster, a SPIFFE-based issuer rotates certificates automatically in hours.

If You Genuinely Must Keep One

Record it in your certificate inventory with an owner. Pin the public key explicitly in the client rather than disabling validation wholesale. Keep the validity period short enough that renewal is a routine, scripted event. Use a 3072-bit RSA or P-256 key, set a meaningful subject alternative name, and store the private key with the same care as any other secret. Secure configuration of this kind is also one of the five Cyber Essentials controls, so the certificates on your internal management interfaces are in scope for assessment.

Our Network Security Services routinely surface forgotten self-signed certificates on management planes, and the clients that have been taught to trust them.

FAQ – Frequently Asked Questions

Is a Self-Signed Certificate Insecure?

The cryptography is identical to any other certificate, so the traffic is genuinely encrypted. What is missing is authentication: nothing proves the server is the one you intended to reach. On a network where an attacker can redirect traffic, encryption without authentication protects very little, because the attacker simply presents their own self-signed certificate instead.

Why Does My Browser Show a Warning?

The browser cannot build a path from the certificate to any root in its trust store, so it reports NET::ERR_CERT_AUTHORITY_INVALID or a similar error. The warning is not about encryption strength; it says the identity is unverified. Adding a permanent exception hides the warning on that one machine while leaving every other client, script, and mobile app failing.

Can I Use a Self-Signed Certificate on an Internal Site?

You can, but a private certificate authority is usually less work overall. One root distributed by Group Policy or MDM makes every internal site trusted, gives you revocation and expiry tracking, and avoids teaching staff to dismiss warnings. Tools such as step-ca or AD CS make this a day’s work, and the result scales to hundreds of services.

How Long Should a Self-Signed Certificate Last?

Keep it short, ideally under a year, and automate renewal. Ten-year certificates feel convenient but they outlive the person who created them, the key size that was adequate when issued, and any record of where the private key is stored. A short lifetime forces the renewal process to exist and be tested, which is the real goal.

Related Terms

You May Be Interested In These

What Is Chargeback Fraud?

Cybersecurity Terms

17 November 2026

What Is Chargeback Fraud?

Read More
What Are Web Skimming and Magecart Attacks?

Cybersecurity Terms

17 November 2026

What Are Web Skimming and Magecart Attacks?

Read More
What Is an Exploit Chain?

Cybersecurity Terms

17 November 2026

What Is an Exploit Chain?

Read More