The delegation scale

The question

Which actions can the system perform alone, which with confirmation, which can it not perform at all?

Why it is a decision

Because the corpus says exactly what happens if you do not make it: “without this scale, the level is in effect decided by development, case by case”.

This is not a passive omission. It is an implicit delegation of positioning to whoever implements the individual feature, who decides on local criteria and without an overall view — and who is not at fault, because nobody gave them a scale.

What the literature says

• The three-step model is already in the corpus, with the example: “consulting the catalogue, yes; modifying an order, maybe; cancelling a contract, probably not”. The criteria for moving up a level are reversibility of the action, economic impact, presence of personal or third-party data.
• The scale has to be written twice, for two different subjects: what our system can do for the user, and what an external agent can do on our service on the user’s behalf. They are two decisions with the same schema and different trade-offs, and on the second the corpus is blunt: “blocking them all is a legitimate choice, doing it without knowing is not”.
• The architectural constraint: “a function reachable only by walking a hierarchy of menus is not delegable”. The scale is not just a policy, it is a requirement on how functions are exposed — and so today delegability is a side effect of the architecture, not a choice.
• Confirmation points on the irreversible and recovery paths with human escalation are the corresponding artefacts, attributed to the Trust Designer.

The cost of either branch

Narrow scale: every confirmation is a point of exit, and the system comes across as less useful than the expectation the corpus records as already observable.

Wide scale: the error becomes ours without being contractualised, and drop-off at confirmation points can no longer be seen — because there are no confirmation points.

How it is measured

Cancellation rate after an automatic action: signals a scale that is too wide. Drop-off at confirmation points and share of users switching automation off after a trial: signal a scale that is too narrow, or badly positioned.

These are three of the four indicators of the trust budget: the delegation scale and the trust budget are measured with the same instrument.

Who decides

Product Manager for the scale, Legal / Compliance for the irreversible, Trust Designer for the confirmation points. If the third is missing, the first two produce a list of prohibitions, not a scale.

By when

Action logs, explicit confirmation on irreversible operations and escalation to a human operator are item 5 of the regulatory perimeter. On delegated payments the PSD3 framework is still being defined: there is no date, but there is a constraint.

Where it falls

The user expects the system to act, Making a traditional product usable by an agent, Trust, AI literacy and calibration, The regulatory perimeter (items 5 and 8).