Longer arguments about automation, agents, and the human work that survives the automation. Each one takes a position that could turn out to be wrong.
Direct answers, with the caveats left in.
Agentic AI changes fintech operations by taking over the sequence of lookups between systems, not the decisions at the end of them. The interesting shift is not that a model can reason; it is that a model can carry a task across four dashboards that a human used to open one at a time.
Automation that removes a person from a loop tends to be slower and worse than automation that removes the typing, because the person was doing two jobs: the work and the noticing. You can automate the first fairly easily. The second one you have to design in, or it disappears.