Two shops open without a security warning in your browser. One is a business you have used for years. The other has arrived in an unsolicited message and offers an unusually good deal. Does the fact that both use HTTPS make them equally trustworthy?
No. A website certificate helps your browser check a connection to a particular name. It is not a recommendation to buy from the person behind that name. To see the difference, it helps to follow what a certificate actually contains and how it is used. 5 3
An identity document for a website name
Start with the website name. In an address such as https://portal.example.com, portal.example.com is the name of the service the browser is trying to reach. A domain name is an internet name under which an operator can arrange services. The browser needs a way to check that the responding system is entitled to act for the name in the address, rather than accepting whatever system answers. The certificate contributes to that check. 3
For that check, the operator has two mathematically related pieces of information, called a key pair. One is kept secret: the private key. The other can be shared: the public key. A program with the private key can produce a digital signature over particular data. A program with the public key can check that signature against those data. If the signed data are changed, the original signature no longer verifies for them. Knowing the public key should not let someone create valid signatures as though they had the private key. 5 18
Here “signature” means a checkable mathematical result, not a handwriting image. The check concerns the signing key and the signed data. It does not tell the browser that every sentence on the website is true.
There are two separate identity checks. The issuer signs the certificate: the browser checks that signed statement about the website name and public key. During the live connection, the server must separately prove that it has its own private key. Checking the issuer’s statement and checking the responding server are related, but they have different purposes. 5 18
A web certificate is a digital document. It contains, among other information, names it covers, dates of validity, a public key (information used to check a digital signature) and the issuer’s digital signature. The signature allows software to check who issued the certificate and whether the signed information has changed. It is not simply a picture of a seal that a website can paste onto a page. 5
Think of an identity document with a limited purpose. It helps establish a particular identity under particular rules. It does not review the holder’s customer service, financial health or honesty. The analogy is useful as long as we remember that an ordinary domain-validated web certificate concerns control of a domain name, not a general background check on a company. 4
The issuer is called a certificate authority. For example, Let’s Encrypt describes asking the applicant to demonstrate control of the domain. The applicant might have to make a particular item available on the website or add a specified entry to the domain’s published settings. The authority checks the response before issuing a certificate. That is a practical demonstration of control, not an inspection of the applicant’s invoices or accounts. 4
Why does that demonstrate control? The authority chooses data for the check, and the applicant must arrange for the required response to appear at the specified location. The authority then requests that location and verifies the response. A random visitor who can merely read the homepage cannot normally change its contents or the domain’s published settings to complete the check. This is evidence of control over the relevant name under the issuing method. It does not establish how the applicant acquired that control, nor does it assess every business action taken afterwards. 4
This explains why a misleading website can have a valid certificate. Its operator may really control the name in the address. The problem is that the visitor mistakes that name for another business. A certificate check does not read the visitor’s intentions or decide whether a familiar logo is being used honestly. 3 4
The public part and the secret part
The word “key” causes understandable confusion here. The certificate contains a public key, which is meant to be shared. A corresponding private key must remain protected. In a typical certificate-based TLS connection, the server uses the private key to prove that it holds the secret associated with the certificate. The browser checks that proof using public information. Simply copying the public certificate does not provide that secret. 18 5
You therefore do not need to panic because someone found your website’s certificate on the internet. Sharing it is part of its job. Sharing its private key would be a different matter. Likewise, a customer asking an operator to explain a certificate problem does not need the private key by email. The website name and the reported warning are a much safer starting point. 18 5
The browser also needs a reason to trust the issuer. It checks a chain of signed certificates leading to an authority it already trusts, subject to the relevant validation rules. A certificate cannot make itself trustworthy merely by saying “trusted.” This is why the existence of a certificate file and a successful browser check are different things. 5
For example, the website sends its certificate and supporting issuer certificates. The browser checks the signature on the website certificate using the issuer’s public key, then checks the supporting links back to an authority accepted by its trust setup. That starting authority is not trusted merely because the website sent its name. It must already be an accepted starting point for the browser. The linked certificates are called a chain of trust. A forged certificate can claim any issuer in its text, but it cannot pass these checks just by making that claim. 5
The browser combines that chain check with the intended website name, validity dates and other applicable conditions. During the connection, it checks the server’s proof that it holds the private key corresponding to the public key in the website certificate. Together, those checks connect an accepted issuer’s statement to the system that is responding now. A stored certificate file alone provides only part of this picture. 3 5 18
Which name is actually covered?
Suppose your staff should use portal.example.com. The browser needs a certificate that covers that name. A certificate for some other name is not sufficient just because the same company operates both. Certificates can list several names, but the relevant name must fit the matching rules. 3
There are also certificates with a star in a name, such as *.example.com. Under the service-name rules, that star can replace one part of the name. It can cover portal.example.com, but not example.com itself or accounts.portal.example.com. Those names could be covered separately; the star alone does not cover them. This is why “we have a certificate for the whole company” is too vague. 3
The useful question is whether the certificate fits the exact address your people and software use. The browser checks the certificate’s list of covered names, rather than assuming that one certificate covers everything run by the same business. 3
Dates matter too. A certificate has a start and an end to its validity. But a date in the future is only one condition; the trusted issuer, intended name and other validation conditions still matter. An unexpired certificate is not automatically acceptable for every connection. 5
What to do with a warning or an unfamiliar certificate
If a browser warns you about a certificate, do not dismiss it because the page looks familiar. Report the exact address and the warning through an established support route. The operator needs to identify the problem. A warning alone does not tell you whether the cause is a deployment mistake, a trust problem or something more serious.
There is another source of confusion: a certificate found in a public record is not necessarily the certificate a website is serving now. Public certificate logs record certificate-related information. They do not continuously check that every named website is running and using that certificate. A record about issuance and a live connection answer different questions. 8
What does this mean when choosing a supplier? A properly checked certificate contributes to an authenticated, protected connection. It does not approve an invoice, prove that customer permissions are correct or promise that a company will honour its contract. Those conclusions require different information. 5 18
The next time someone says “the site has a certificate,” ask what they mean: a certificate was found, a browser accepted the intended address, or a company was assessed? Only the first two are certificate observations. The third is a much broader judgement.
Sources
[3] IETF RFC 9525 — Service Identity in TLS
[4] Let’s Encrypt — How It Works
[5] IETF RFC 5280 — X.509 Certificates and Revocation
[8] Certificate Transparency — How CT Works
[18] IETF RFC 9846 — TLS 1.3 (July 2026; obsoletes RFC 8446)