What happened. Major banks are assigning AI agents system logins, recurring tasks and human managers as they expand agent use across client onboarding, wealth management, trading and internal operations.
What it actually means. The operating-model conversation has arrived before most companies have finished the software conversation.
Once an agent has a login, daily responsibilities and somebody accountable for its performance, it is no longer being treated like a feature. It is being managed like a worker. That creates new questions around ownership, access, training, quality control and what happens when an agent’s performance deteriorates.
Companies are not just buying AI. They are beginning to manage mixed human-and-agent teams.
What to watch. Whether formal agent-management responsibilities begin appearing inside operations, IT and procurement job descriptions.