A more useful/practical question would be: what tools can we build to assist CEOs in their decision making and make them more productive? I think the answer will come out as "not much", since the job is so high level and abstract. You'd have a better time targeting lower-skilled jobs or technical jobs like software engineering.
They are called data scientists. An AI would be able to crunch numbers, but someone needs to read and interpret the results. You can't automate that part yet, humans are still the best in interpreting data into social and business context.
Hire people, fire people, pitch investors, manage direct reports across multiple business lines, understand what's going on in the world and market, set a product vision/direction, shape the culture of the org, etc etc. That's AGI territory.
We may be able to formulate our current understanding of what it takes to be a CEO in every feasible future context that we can imagine, and create an agent (not an AGI) that does that.
But then what would happen when the distributions change out of sample in an unforseeable way? Suppose two countries go to war and this drastically changes the operating environment of the business. The agent would need to learn how to operate with human-level capability in this novel environment which it wasn't trained specifically to do. That's why I'm thinking it requires an AGI.