When a password is genuinely worth adding
Encryption earns its place when a document will travel somewhere you do not control. An email attachment passes through several servers and sits in an inbox indefinitely. A file shared into a group chat can be forwarded by anyone in it. A document left on a shared drive is readable by everyone with access to that drive, which is usually more people than intended.
In those situations a password means that possession of the file is not the same as access to its contents. That is a meaningful improvement for salary details, medical letters, identity documents, contracts and anything covered by a confidentiality obligation.
It is worth being clear about what it does not fix. Encryption protects the file in transit and at rest; it does nothing once the recipient has opened it. If they save an unlocked copy, forward it, or print it, the protection ends there. Encryption is a control on distribution, not a control on the recipient.
Choosing a password people can actually use
The hard part of encrypting a document is not the encryption; it is getting the password to the recipient without undermining the point. Sending the file and the password in the same email is the most common mistake, and it reduces the whole exercise to theatre — anyone who reads the message has both halves.
Send them by different routes. Email the document and give the password by phone, in person, or through a messaging app the recipient already uses. If that is impractical, agree a scheme in advance — a shared reference number, a known date — so no message ever contains both.
Make the password long rather than clever. A passphrase of several unrelated words is far harder to attack and far easier to read out over a phone line than a short string of substituted characters. And write it down somewhere you will find it: a document you have encrypted and cannot open is functionally a document you have destroyed, and there is no recovery path.
What the encryption actually protects against
Modern PDF encryption using AES-256 is strong. Against a well-chosen password there is no practical way to read the contents without it, and the protection is real rather than nominal.
The weak link is almost always the password itself. A common word, a name, a date of birth or a short numeric string can be attacked by a program that tries millions of candidates, and the encryption strength is irrelevant if the password is guessable in an afternoon. Length is the defence that matters most.
It is also worth distinguishing this from permissions-only protection, which some tools present as equivalent. Marking a document as not-printable or not-copyable relies on the reader software choosing to honour those flags, and plenty of software simply does not. A permissions flag is a request. A password is a lock.
Encryption compared with the alternatives
A password on the file is not the only way to control who reads a document, and it is not always the best. A shared drive with proper access controls gives you revocation, which encryption cannot: once someone has the file and the password, you cannot take it back. A link with an expiry date achieves something similar for one-off sharing.
What encryption gives you that those do not is independence from any service. The protection travels with the file, works offline, needs no account on either side, and does not stop working when a subscription lapses or a company changes its product. For documents that will be archived for years, that durability matters.
The pragmatic answer for most people is to use both: share through a controlled channel where one exists, and encrypt the file anyway for anything sensitive enough that a misdirected email would be a genuine problem. The two protections fail in different ways, which is precisely why they combine well.