The reason open code is trusted by people who never read it

A protocol’s contracts were open-source, and users cited this constantly as the reason they felt safe. A survey of those users found that fewer than one in fifty had opened the repository, and none had read it line by line. The code being public did not mean it had been checked; it meant it could be. The safety people felt came from the availability of inspection, which they mistook for the fact of inspection, by themselves or by anyone they could name.

What everyone sees

A user hears open-source and imports the belief that many eyes have reviewed it and would have caught anything wrong. The phrase carries the reputation of collaborative scrutiny. The individual user, who has not read a line, borrows confidence from an imagined crowd of reviewers, assuming that someone qualified must have looked because the door was open. The possibility of review is felt as the guarantee of review.

What is actually happening

Security researchers call this the many-eyes fallacy: openness enables review but does not perform it, and studies of open-source projects show that most code, especially in smaller or newer projects, is read by almost no one. Serious vulnerabilities have persisted for years in public repositories precisely because availability was mistaken for audit. Trust that rests on transparency requires someone to actually exercise the transparency; when no one does, the code is effectively unreviewed while enjoying the reputation of being reviewed.

Why it stays hidden

The hidden mechanism is the transfer of trust to an absent reviewer. Each user assumes the checking has been done by others, and every other user assumes the same, so the review that everyone relies on is performed by no one. Openness supplies the alibi for this diffusion of responsibility: because inspection was possible, its absence is never noticed, and the code inherits the credibility of scrutiny it never received.