Thinking this through a bit more: As I understand it, many published ML papers detail how to train a model, how to predict from the model, and sometimes code that represents a model that has been trained on data that can take new data and predict on it.

I wonder if there is an example of someone objecting to a paper that did not release the training data or software that can predict (aka just the algorithms).

The point I'm trying to get at is: it seems that most ethical objections are to the result of applying algorithms to ("bad") data, not the algorithms themselves.