What “reports to” should actually mean

"Reports to the Chief Product Officer" and "has a reporting line that actually protects independence" get treated as the same statement on an org chart. They are not, and the gap between them is the missed value work.

The org chart problem

A reporting line drawn for administrative convenience, who does the performance review, who approves leave, often ends up being the same person or team whose work the individual is meant to 'assess.' That is not a hypothetical conflict, it is the default outcome of treating this like any other functional hire rather than naming the conflict up front and designing around it.

What a protective reporting line actually requires

Separation between who evaluates the job role's performance and who evaluates the AI ethics of the organisation/board/work. A named recipient for findings who sits outside the team under review. And a way for a serious, unresolved finding to reach that recipient even if the ethicist's immediate manager would rather it did not, which is the mechanic our piece on negotiating independence clauses covers in contract terms.

Dotted lines, dual reporting, and their limits

A dotted line to a board committee or an audit function is a reasonable structural fix, but only if it is a real channel someone actually uses, not a line on a chart that has never once carried a message. Test it before relying on it: has anything ever actually travelled up that dotted line, or does it exist only on paper...

The test: who can silence a finding

Ask plainly, before the job is filled: if this role produces a finding the direct manager disagrees with, who has the authority to stop it going further, and does that person have any stake in the outcome being suppressed. If the answer to the second half is yes, the reporting line has not actually solved the problem, whatever the org chart says.