Are password managers actually safe to use?

The concern people raise is reasonable: putting every credential in one place seems to create a single point of failure. In practice a password manager is a very large improvement over the realistic alternative, which is reused or weak passwords across dozens of accounts.

The reason it is safe rests on how they are built. Reputable managers encrypt your vault on your device before anything is transmitted, using a key derived from your master password. The provider stores only encrypted data and never receives the master password, which is what people mean by zero knowledge. A breach of the provider therefore exposes encrypted blobs rather than usable credentials, and that has been borne out in real incidents.

That design has one important consequence: if you forget the master password, nobody can recover it for you. There is no reset, because there is no copy. Recording it somewhere physically secure is a genuine requirement rather than optional caution.

The realistic threats are worth naming honestly. A weak master password undermines the whole model, because it is the only thing protecting the vault, and older vaults encrypted with outdated settings have been shown to be more attackable than expected. Malware on your own device can capture credentials as you use them regardless of where they are stored. And phishing still works, though good managers help by refusing to autofill on a domain that does not match.

The practical guidance is a long unique master password, two-factor authentication on the vault itself, and keeping the software current. Compared to the risk of reuse, where one breached site compromises every account sharing that password, the trade is strongly favourable.