Sure, but that's the same as "if I didn't torture them, if I didn't slaughter them, someone else would have," which has failed to be enough defense to stop war criminals from being hanged.
Regardless of what comes after, you're never wrong to do the right thing.
That's one of the strange problems here. It'll probabally get farmed out to a less technically proficient team which, ironically, could lead to an increase in the loss of innocent life.
That could be the outcome, yes, or the outcome could be entirely different. It could be that the projects gets farmed out to a less-than-competent contractor that is willing to work on it, with a result that is so spectacularly bad (along any given dimension, not just how many innocent lives are affected) that the project dies.
That, too, is a complete guess and just one of many hypothetical outcomes. I don't think there is any sufficient ground to infer that "if Google does not do it, it will be worse".
I think you're falling into the perfect market fallacy. Quality AI work is not a commodity. It's entirely possible they either won't be able to find any credible bids or they'll find a bid from a less competent party and the project will have to be scaled back because they can't accomplish all their goals well enough.
Maybe another shop can do it all the same, the important matter is that such a prestigious workforce as makes up Google has rejected this kind of work. Money cant buy belief in armistice that quality of publicity.
