We are witnessing a revolutionary shift from basic AI chatbots to true cognitive agents — systems that can think strategically, plan, and learn from their successes and failures. However, if we always put humans in the loop, we are unlikely to attain the true benefits of AI transformation. So, how much freedom should we grant AI agents? Surprisingly, the answer is about more than how big the risks are; it’s about how well we understand those risks. Too little latitude, and the agent’s capabilities are constrained to answering basic questions. Too much autonomy, and brand, reputation, customer relationships, and even financial stability are at risk. One approach is to wait until a general regulatory and commercial consensus on agentic AI emerges. Alternatively, for those bold enough, uncertainty can be harnessed as a decision rule for knowing what to do next.
