AI does not think.
It predicts from what it was fed.
Every institution deploying it
has been asking the same question.
Who owns the decision the model makes.
The question is wrong.
The model does not make one decision.
It makes ten.
Nine of them happened
before anyone in your building
looked at a screen.
Someone decided what data the model
would be allowed to see.
Someone decided what it would optimize for.
Someone set the number
where a flag fires
and where it stays silent.
Someone decided whether the output
waits for a person
or moves on its own.
None of those people
are in the room
when the call gets made.
The one who is in the room
signs for all of it.
What most institutions track.
↳ Whether the model is accurate
↳ Whether the vendor is credentialed
↳ Whether a human stayed in the loop
What the record actually holds.
↳ One name
↳ One timestamp
↳ Nothing about the nine decisions
that shaped the call
In Clinical AI the chart records the clinician.
In Financial AI the log records the trade.
Neither one records
what shaped them.
That is not a documentation problem.
It is ten decisions
wearing one signature.
You already know which of the ten
has no name on it
in your institution right now.
Which one gets named first?
Mo Johnson, MD MBA is a cardiothoracic surgeon and the founder of GPe Research. He created The Accountability Gap™ (TAG™), which names the moment the AI stops and the named owner starts. MedicoVigilance™ covers clinical AI. FinVigilance™ covers financial AI. Field Notes are short dispatches from the accountability frontier, published at tagframework.com.
Follow the work on LinkedIn: linkedin.com/in/mo-johnson