As it's often used, the anthropic principle is fatally flawed.

It often starts with an argument between a creationist (could also be an advocate of intelligent design, but I'll just call them the creationist) and an evolutionist. The creationist says, look, the origin of life by purely naturalistic means is ridiculously improbable (and therefore it's reasonable to consider the possibility that God did it). They trot out some generally-accepted scientific principle, do a back-of-the-envelope calculation, and come up with a number that is, in fact, ridiculously improbable.

The evolutionist responds with the anthropic principle - if no life had arisen in this universe, we would not be here arguing about how life arose. This is clearly logically correct. It is also completely irrelevant.

The creationist didn't argue that life couldn't have arisen in this universe. They argued that it could not have arisen by purely naturalistic means. They're arguing about how, not about whether. The creationist might answer: "Yes, I agree that if life had not arisen in this universe, either by creation or by naturalistic processes, then we would not be here having this conversation. But the question is, which way did life begin?" The anthropic principle doesn't address that issue whatsoever.

It doesn't address that issue unless you add an assumption - that life had to begin by purely naturalistic means, that is, that the probability of creation is precisely zero. Then the anthropic principle is relevant, but then there's a new issue, that of begging the question.

I suspect that this assumption is present on the side of everyone on the evolution side that pulls out the anthropic principle in an argument with a creationist, but I have never heard it explicitly stated. I'm not even sure the evolutionist realizes they're making an assumption - it's so ingrained in their world view that they can't think that the alternative might be possible.

I grant you that the universe is huge and the evidence available is small. But that can turn into an "evolution of the gaps" argument quite easily, so I'm not sure you want to seriously use it.