Computer passwords began as a way to share a machine.

A mainframe needed a way to tell its users apart

In 1961, MIT’s Compatible Time-Sharing System let multiple people use one computer by taking turns on its processor so quickly that each person could work as if the machine were theirs. But sharing the computer raised a practical problem: how could it keep one user’s files separate from another’s?

Fernando Corbató’s answer was to require users to sign in with a password. The system could use that secret to identify an account and limit access to its files. A password did not make a file unreadable by encrypting it; it acted more like a key at the door, helping the system decide who was allowed in.

One simple key became many

On a shared mainframe, a password tied to a single account could solve a clear, local problem. As computing spread across more services and devices, people encountered the same basic mechanism again and again: another account, another login, another secret to remember or recover.

That multiplication changed the tradeoff. A password is useful only if the person who owns the account can recall it and keep it from others. Users may reuse or simplify passwords to cope, while systems must handle forgotten credentials and attempts to guess or steal them. The original tool for separating files can therefore become a security burden of its own.

Corbató later called the invention a mistake—not because the shared computer no longer needed a way to distinguish users, but because the password solution scaled into a daily chore. The irony is that a safeguard designed for one machine’s file boundaries became a ritual repeated across much of digital life.