Let's not confuse an ethics team with HR. The later exists to ensure proper legal adherence with a varying degree of benefit to the employees. The former[1] is an electively instated position that usually has a dual mandate somewhere between the polls of "Make leadership confident that teams don't make them lose sleep" and "Make teams confident that leadership doesn't make them lose sleep" (usually leaning strongly towards the second) as a sort of employee moral assurance team. I don't disagree that the ethicist is partially there to justify actions but they're mainly there to be seen as an obstruction to particularly egregious actions so that, if, for instance, you have a boring data center maintenance job, you can have the belief that nobody is using the product of your labor to for something heinous. The efficacy of such teams is quite questionable but there is some logic in assigning the job of worrying about ethical considerations to one person that can shove their nose in every crack and cranny so that every individual employees can talk to/lean on that person instead of needing to go to such lengths themselves if they feel compelled to do so.

Most people don't like to do evil (even if the money is good) so allowing an economy of scale around morale reassurance is rational. This obviously breaks down if the company is seen to hide things from that department or if that department is clearly, themselves, unethical. The first of those conditions are why ethics departments at Meta and OpenAI have very little efficacy.

1. Outside of extreme circumstances where it has been mandated to be a specific regulatory board because of past actions or fear of future actions in which case it's better viewed as part of the bureaucracy of the governing body (usually the state, sometimes a shareholder or other concerned party).