Are password managers safe? An honest answer

Password managers are safer than what almost everyone does instead: reusing a handful of memorable passwords across dozens of accounts. That much is a genuine yes. But “safer than the alternative” is not the same as “safe”, and the honest version of this answer needs both halves.

The risk you already have

Before judging a password manager, price the status quo. If you reuse one password across ten sites, your security is set by the worst-run of those ten. When that site leaks (and over the course of a decade, one of them will), attackers take the email-and-password pair and replay it automatically against every major service within hours. This is credential stuffing, and it is the attack that actually reaches ordinary people. It requires no cracking, no targeting, and no skill.

A password manager eliminates that attack completely, because there is no reusable pair. That single change is worth more than every other password decision you will make.

What actually goes wrong with password managers

Three things, in descending order of how often they happen.

Your master password was weak. This is by far the most common failure, and it is the one you control. Every vaulted manager’s security collapses to the strength of one password. If yours is a word with a number after it, encryption buys you very little.

The provider was breached. It happens. In 2022, LastPass suffered a breach in which attackers obtained customer vault backups. The vaults were encrypted, but encryption is a delay, not a wall: attackers get unlimited offline attempts, and accounts still using older, low-iteration key derivation settings were meaningfully exposed. Investigators later linked the theft of more than $150 million in cryptocurrency to those stolen vaults. The lesson is not “never use a password manager”. It is that a stored vault is a permanent, patient target.

Your own device was compromised. Malware with access to your machine can capture your master password as you type it, or read decrypted contents from memory. No password manager architecture defends against an attacker who is already inside the device.

Notice what is missing from this list: someone breaking the encryption itself. That is not how these incidents go, and designing your habits around it wastes attention that belongs on the three above.

The concentration question

The reasonable objection to password managers is not “is the crypto good”. It is: should everything I own live behind one thing?

That concern is legitimate, and it has two different answers depending on the architecture.

Vaulted and deterministic password manager trade-offs

Question	Vaulted manager	Deterministic manager
What is stored	An encrypted copy of every password	Non-secret account metadata and security records, not generated credentials
What a breach yields	An encrypted vault, attackable offline forever	No stored list of generated credentials, but backend metadata still matters
Recovery if you forget	Often possible via recovery codes or emergency access	Impossible by design
Autofill	Yes	No, copy and paste

A vaulted manager concentrates risk and pays you back in convenience: autofill, secure notes, recovery, sharing. A deterministic manager avoids storing generated credentials, and charges you for it in convenience: no import, no autofill, no recovery.

Neither is the right answer for everyone. Both are dramatically better than reuse. If the mechanics behind that table are new to you, how do password managers work? walks through both designs step by step.

How to judge whether a specific manager is safe

Five questions, in order of how much they tell you:

1. Is it zero-knowledge? Can the provider reset your password and still give you your data? If yes, they hold a key, whatever the marketing says. See zero-knowledge encryption.
2. What is the key derivation setting? OWASP currently recommends at least 600,000 PBKDF2 iterations, or a memory-hard function like Argon2id. Providers that quietly left old accounts on 5,000 iterations are the reason the LastPass breach hurt as much as it did. Why iteration counts matter.
3. Has the code been audited, and was the report published? “Audited” without a public report is a claim, not evidence.
4. What is the breach history, and how did they communicate? Every long-lived provider has incidents. What separates them is whether the disclosure was prompt and complete.
5. Can you get your data out? An export path you have tested is the difference between switching providers and being stuck with one.

What to do this week

• Turn on two-factor authentication for your email account first. Email resets everything else, which makes it the highest-value target you own.
• Make your master password a six-word random passphrase rather than a clever sentence. Six random words is roughly 62 bits; a sentence you invented is often under 30.
• Stop reusing passwords on the accounts that matter: email, banking, and anything holding payment details. You do not have to fix all hundred at once.
• Check whether your manager’s iteration count is current. On several providers this is a setting that was never upgraded on old accounts.

In short

Password managers are safe enough that the security profession uses them, and imperfect enough that the profession also argues about which kind. The failure modes are known: weak master passwords, stored vaults that leak, and compromised devices. Pick an architecture whose trade-offs you can live with, make the one password you must remember genuinely strong, and stop reusing.

If the stored-vault half of the risk is what bothers you, MoolKey removes it by not storing anything, and is equally clear about what that costs you.