The team works, the tools are available and, even so, every deliverable comes back to the same person. The leader looks for context, answers questions and confirms details. Before automating that flow, it is worth understanding why so many handoffs depend on them — knowledge that was never written down, unclear responsibilities or decisions that really do need their involvement.
Separate information, criteria and decisions
Look at one recurring responsibility and the questions that reach you. Some only ask for a known piece of information. Others require applying a criterion. Others are new choices, with consequences that need your judgment.
A director who reviews every proposal from a team usually sees the same corrections: an outdated description, a missing condition, the wrong format. Those criteria can be written down and used in a preparation step with AI. A negotiation that changes the company's commitment, on the other hand, still needs the director's decision.
Test a preparation step that reduces ambiguity
AI can gather information, compare the deliverable with the written criteria and highlight exceptions. The first test can produce a list for review, with sources and with what it could not resolve.
Check whether it helps. Maybe people start sending more complete requests. Maybe conflicting criteria come to light. Maybe it becomes clear that the problem is an approval rule that is too centralized.
After a few uses, look at what still reaches you and why. If the exceptions are still concentrated and the review only grew, the design deserves another look.
In AI Thinking by Dani Hayoshi, technology takes part in the redesign, but the change depends on agreements about responsibility and decisions. The goal is to make the work understandable and easy to distribute. See also AI for leaders.
Dani Hayoshi is the author of AI Thinking. Her work connects artificial intelligence, education and entrepreneurship. About Dani.