In much of mainstream society, it is considered kind to "put an animal out of its misery" if it is suffering and will likely not recover. Yet it's considered morally abhorrant to do the same to a human.

It surprised me to learn that, despite their willingness in general to break with conventional morality, much of EA appears to feel similarly. (Example, and another example. Content warnings for animal death, etc.)

Yet I haven't been able to find any good argument for this. If your goal is to reduce suffering, then killing humans in pain acomplishes that just as effectivly as killing animals in pain does. If your goal is to maximize preferences satisfied, I'm pretty sure those animals would prefer to not die.

Obviously you can justify this by just saying "my utility function sees animal death as neutral and human death as negative", which is a perfectly valid utility function for someone to have, but EA usually tries to provide more objective grounding for their morality.