From rules posted in advance to a decision backed by evidence
The model starts before anyone acts on a critical request. First the posted rules, then the seal, then who it was for, then the examination, then the business decision. Each step answers one question and gives the next step a fact to work from, not a guess.
How it works
The steps, in order
Start with the rules posted in advance
The company posts which kinds of request it says it always seals. The person receiving a request should not have to trust the request itself to find out what the sender has committed to, so the rules are readable on their own.
Seal the request, and show who sealed it
When a covered request is sent, the company creates evidence tied to that exact request. The point is to show which organisation stands behind the seal, not to claim that everything which follows has already been proved.
Work out who it was for
A different question from who sent it. Who the request was for is part of the evidence, not just an address that appears in a message, so a genuine request cannot be reused by anyone who simply gets a copy of it.
Examine the evidence, step by step
The checker does not boil this down to a green-or-red judgement. It reports what the evidence proves at each step, and never claims more than the proof underneath supports.
Use a phone line that is genuinely independent
Some situations still need a person to confirm by phone. An independent line is a number you already had, never a number given in the request itself. If the message says “call us on this number”, calling it proves nothing: the attacker may control that number too. Use a number from your existing supplier records, or a contact you had before the request.
The decision stays a business decision
Atestan does not move legal responsibility onto the system, and attaching evidence to a request does not make a company compliant with anything. A person still makes the call. Atestan just gives them better evidence to make it with.
The spoken code reveals nothing by itself
Over the telephone, the caller dictates the code and names ONE field to confirm — an account ending, a date, an amount. What travels is a code, not the content — and the answer is: matches, or does not. The code discloses nothing on its own: it is a confirmation mechanism, not a secret that proves who is speaking. Built and tested; the service is running, and no organisation is enrolled, so there is no seal to confirm against today.