The objection is simple and banal: Utility contains terms for things like "it's bad to give in to blackmail, as this leads to more expected blackmail in the future*" - consequentialism doesn't have to be short-horizon, blind and dumb. You assess all consequences of an act.

My personal objection (why I'm consequentialist but not utilitarian as usually defined): Caring about others' utilities is not something I have to do because of some Rawlsian argument; it's just something that's already in my utility function because that's how my brain evolved to be. You can do approximations that are equivalent to "weighting people's utilities" based on your thoughts, feelings, whims, their likeability, the uncertainty you have about them, etc. And those weights can be negative, because why not? Spite is also natural. If someone tries to threaten his own bodily integrity, see if I care.

^(*: even accounting for all the not-cut fingers, and for everyone's utilities and not just yours, the "giving in to lots of blackmail" future is worse than one where you don't, which does need to be argued for but isn't hard to argue. As opposed to e.g. "giving in" to win/win trades.)