AI can do the task. Who owns the decision?
I recently watched an AI assistant handle one of the smallest annoyances in business. An email came in. The assistant responded. It sent a scheduling link. The other person picked a time. I did nothing. Then a meeting I actually wanted appeared on my calendar. Beautiful. That is exactly the kind of work I want a machine to take away: repetitive, predictable, useful, and not especially improved by my personal involvement. But the assistant did not take the meeting for me. More importantly, it did not decide what the meeting meant. Once the call started, another human was there. Maybe several. Tone mattered. Trust mattered. Priorities changed as we talked. Someone asked a question nobody had planned for. We had to decide whether there was anything worth doing next. The task was automated. The judgment was not. The same thing is starting to happen everywhere. An AI agent can recommend the next account to call, the next product to offer, or the next action to take. Often that is incredibly useful. But if customers are not choosing you, automating more activity does not necessarily solve the problem. A machine can tell you whom to contact. It cannot guarantee that the choice you are asking another person to make makes sense in that person’s world. Of course, we already hand decisions to machines when the choice is boring enough. Toilet paper is a good example. If I know the brand, quantity, and how often the household needs it, put it on auto-renewal and spare me the ceremony. I do not need a fresh strategic review every time the garage gets down to the last twelve rolls. Even then, I keep the right to stop it, speed it up, slow it down, or decide that somehow we now own enough toilet paper to survive a small siege. That suggests a useful dividing line. The more repetitive, predictable, and reversible the decision, the easier it is to delegate. The more ambiguous, consequential, or human the situation becomes, the more judgment matters. The interesting question for companies adopting AI agents is not simply: What can we automate? It is: What work should the machine do, and where does a human still need to understand the decision? AI can remove enormous amounts of work. It does not remove the need for a customer to decide. And that is still the part that determines whether they choose you.