Theory
Credentials as insurance
Training, credential, licence, supervision, and liability cover exist as a stack because each layer is insufficient alone. An AI worker arrives with none of them.
The stack
Each layer exists because the one below it does not settle the question of who is accountable when the work goes wrong.
Why it was built this way
Training says a skill was acquired. A credential says someone independent attested to it. A licence says the state permits the act. Supervision says someone reviews the work in practice. Liability cover says there is a payer when all of that fails.
None of these layers is about capability alone. They are about who bears the cost of an error, which is the question an organization actually has to answer before it lets anything act on its behalf.
The hole
There is no credential for an AI worker, no established insurance product covering its errors as a class, and little precedent to price the risk against. The gap is not technical and it will not close because the models improve.
The practical consequence for Zero Neo is that accountability stays with a named person until the layers exist. That is a constraint on how far the assembly can go, and it belongs in the cost comparison rather than in a footnote.