You seem confused. Read the article you linked. Or, in case it wasn't clear enough: AI must serve humanity. So must the church.

For reference, the general form (so that I can use letters to refer to parties and avoid convoluted phrasing):

(Axiom: B has done X.)

A (to B): You have done X, and you should not have done so.

C: I note here that certain prior actions of A could also reasonably be characterized as X.

D (to C): Ah, here you commit the fallacy of "tu quoque".

This argument is not sound. It misunderstands the fallacy. (To be clear: Wikipedia describes the fallacy accurately; it's just that it's rare in practice, and very often falsely accused.)

Everyone should uphold the standards to which they hold others (and I consider it an obvious moral failing not to do so). The fallacy only applies where C either continues on to argue that B has, somehow, not actually done X (because A did); or, at least, clearly has the purpose of distracting from the fact that B has done X.

But there is nothing fallacious about simply pointing out that A does not live up to A's own implied standards. There is nothing fallacious about the implication that A is therefore being either i) dishonest about the anti-X belief, or ii) simply hypocritical. (To be fair, we don't know, from the given information, which of those is the case; but I think it's fair to say that neither is "fair dealing" and that A is thus a legitimate target of criticism regardless.)

It is also not fallacious for C to use this as a jumping-off point to argue that X is in fact okay to do, although of course this requires further support.

[flagged]