If you are told an organisation has a Chief AI Officer, the title tells you less than you might think. It can be a distinct seat, or a second hat on someone who already has a full job. What tells you more is who is doing the duties.
In May 2024, NASA named its first Chief AI Officer. The release described the role as an expansion of its chief data officer's role. It was the same seat with a second hat on it.
That fits what US federal policy allows. Both of the OMB memoranda that set out the role, M-24-10 in 2024 and M-25-21 in 2025, let an agency head designate an existing official, a chief information, data, or technology officer for example, as the Chief AI Officer, provided that person has significant expertise in AI.
There is one difference between the two memoranda. The 2024 one required that, at the largest agencies, AI be a primary role of whoever held the title. The 2025 one dropped that condition. NASA's release does not say whether AI became a primary role for its Chief AI Officer.
My view is that accountability attaches to the duties, wherever they sit. If an official with another title does the work, the work is done. If the title exists and nobody does the work, the seat is empty in practice, however good the name plate looks. A title says who is meant to answer. The work says who does.
The fair concern about the second hat is attention. A chief data officer carrying AI as an extension has a day job already, and attention is the obvious thing to worry about. I do not know how that has played out in practice, and I have not seen evidence on it either way.
The ten duties in the 2025 memorandum are a better description of the job than the title is. They are what to ask about when someone tells you a Chief AI Officer exists.
My organisation does not have a named Chief AI Officer yet. A good part of what I do day to day lines up with those duties anyway, which is why the difference between a title and the work is not academic for me.