We are automating our entire customer support queue with an AI agent. How do we structure the customer support seat on the Accountability Chart so a human still has GWC™ for the AI outcomes, especially when the tool is highly autonomous?

Question

Tyler Smith · Accepted Answer

Creating an Accountability Chart that relies on AI requires a strict adherence to the EOS® principle that only humans can occupy seats. You cannot put an AI agent in a seat. A human must always be accountable for the measurables and deliverables of that seat.

For a leader to GWC™ their seat, they must truly understand how to manage the outputs of the AI tools they use. They do not need to know how to write the underlying code, but they must know what a good output looks like and how to spot errors. If a team member does not understand the business logic behind what the AI is producing, they do not GWC™ the seat.

If you have an automated process that nobody on the team fully understands, you have a massive operational risk. You must simplify the process. Either train the seat holder so they have the capacity to audit the AI work, or adjust the seat description.

In your next weekly Level 10 Meeting™, review any seats where AI tools are heavily utilized. Ensure the seat holder can explain the business logic of the tool and is fully prepared to take the hit if the AI produces an error. Accountability cannot be automated.