Computing Library › Security & Safety-Critical Computing
Security & Safety-Critical Computing

Fail-Safe vs Fail-Secure

When a system fails, it can default to the safe state or the secure state, and these are not always the same choice.

What Happens on Failure

Every system will eventually fail, lose power, crash, or lose communication, and the designer must decide what state it defaults to. Fail-safe means it defaults to the state that minimizes harm to people. Fail-secure means it defaults to the state that maximizes protection of assets. The canonical example is a door: fail-safe unlocks so people can escape a fire; fail-secure locks so intruders cannot enter.

The Same Word, Different Fields

Kronos motion — is it safe

Terminology varies. In safety engineering, fail-safe means the system moves to a benign condition. In security, fail-secure (sometimes fail-closed) means access is denied on failure, while fail-open means access is granted. A firewall that fails open lets all traffic through, convenient but insecure; one that fails closed blocks everything, secure but potentially disruptive.

Choosing Correctly

When They Conflict

The door example shows the tension directly: you cannot be both fail-safe and fail-secure with a single mechanism. Real facilities resolve this with layered design, for example emergency egress that releases on a life-safety signal while normal access stays fail-secure otherwise.

Fusion Context

A fusion plant's protective functions are fail-safe by design: on loss of control or power, the machine moves toward a safe, de-energized condition rather than an uncertain one. In the Hyperion breeder and burner designs, a magnet-protection trip or loss of the control link drives the system toward its safe state, because the worst outcome of failing safe is a halted machine, while the worst outcome of failing the other way is equipment damage or a hazard.