Good faith and the record

There is a distinction the law has cared about for a very long time, and that ordinary fairness cares about just as much, though it rarely names it. It is the difference between the person who was careful and got it wrong, and the person who was careless and got it wrong. The outcome can be identical — the same harm, the same loss, the same person on the receiving end — and yet we judge the two cases differently, and we are right to. One was a mistake. The other was a failure of duty. We forgive the first in a way we do not forgive the second, because we hold that doing your honest best and falling short is not the same moral act as not bothering.

This distinction is doing enormous work, quietly, across almost every domain where decisions carry consequences. Negligence law turns on it. Professional standards turn on it. So does the everyday moral sense by which we decide whether to extend someone a second chance. We do not actually demand that institutions never err. What we demand is that they err in good faith — that the mistakes be the residue of genuine care rather than the symptom of indifference. The whole apparatus of accountability is built to separate those two, and to treat them as the different things they are.

But here is the part we rarely say out loud: the distinction is invisible from the outcome alone. You cannot tell, from the harm, whether it came from care or carelessness. A careful decision and a reckless one can produce exactly the same wrong result. The only place the difference lives is in how the decision was made — what evidence was consulted, what rules were followed, what was weighed before the conclusion landed. Good faith is not a property of the outcome. It is a property of the process, and the process is precisely the thing that vanishes the moment the decision is over.

Without a record, every error looks the same

This is the trap that opens when a decision leaves no account of itself. Strip away the record of how a decision was made, and the careful error and the negligent error become indistinguishable. They present identically: a bad outcome, and nothing behind it. There is no way, from the wreckage, to reconstruct whether the institution did its honest best or simply did not look. The distinction the law spent centuries refining collapses, not because anyone repealed it, but because the evidence it depends on was never kept.

And this hurts in both directions, which is what makes it so corrosive. The institution that acted in good faith — that consulted the right evidence, followed its rules, weighed the matter properly, and was unlucky — has no way to prove it. Its care left no trace, so its care cannot be shown, and it stands accused on exactly the same footing as the one that was reckless. Meanwhile the institution that acted carelessly has nowhere it can be caught. Its carelessness also left no trace, so it can claim good faith with total impunity, and there is nothing to contradict the claim. The recordless world is the one where the careful cannot clear themselves and the reckless cannot be convicted.

Good faith is not a property of the outcome. It lives in how the decision was made — and when that vanishes unrecorded, the careful and the reckless become impossible to tell apart.

The record is what makes good faith legible

A decision record changes this, and it does so with a kind of moral precision worth dwelling on. By preserving how the decision was actually made — the evidence consulted, the rules active at the time, the basis on which it was reached — the record makes good faith legible. It converts care from an invisible internal state into something external and checkable. The institution that did its honest best can now show that it did, because the record is the showing. And the institution that did not can no longer hide in the ambiguity, because the record reveals the absence of care as plainly as it reveals its presence.

Notice what this does to the two parties. It protects the careful. An institution acting in good faith should want a record more than anyone, because the record is the only thing that can vindicate it when an honest decision goes wrong. Without it, good faith is just an unprovable assertion. With it, good faith becomes a demonstrable fact — the evidence was here, the rules were these, the reasoning ran thus, and the outcome, though unfortunate, followed from genuine diligence. A Decision Receipt is, in this sense, the careful actor's best defense: it is how diligence proves itself after the fact.

And it exposes the reckless. The same record that vindicates care indicts its absence. The decision made without consulting the evidence, or in defiance of the rules, or with no weighing at all, cannot dress itself up as good faith once the record shows what actually happened. The contestability the record provides is what lets the distinction the law has always cared about finally apply to automated decisions — to separate the forgivable error from the negligent one on the basis of evidence rather than assertion. This is not surveillance of the institution; it is the thing that makes the institution's honesty knowable, which is what good faith was always supposed to mean.

So the record is not the enemy of the institution acting in good faith. It is that institution's strongest ally, and the careless actor's only real constraint. It restores, to a world of decisions that happen too fast and too silently to witness, the oldest distinction in fairness: that an honest mistake and a negligent one are not the same, and that we are entitled to tell which is which. Without the record, every error looks alike. With it, good faith becomes legible again — which protects the careful, exposes the reckless, and lets accountability mean what it has always meant.

— Dispatches · Summit Cognitive