It's rather subtle trying to decide what counts as gender preferential treatment. For example, suppose the hiring committee decides to interview Bob, Carl, and Dave. As a sanity check, someone goes through the applications from women to see whether anyone was overlooked, and they are impressed by Alice's application. There's some debate about whether she looks quite as strong on paper as the other three, but the department decides to interview her as well. Alice is extraordinarily impressive in person, and once all the interviews are complete and the department has learned more about her work, she is the unanimous first choice. Does this count as preferential treatment? A male applicant might not have been rescued from being overlooked the way Alice was, but he might have been less likely to be overlooked in the first place, so it's difficult to give an objective answer (it depends on which counterfactual scenario you imagine). Gender was not relevant for the decision once all the information was gathered, but deliberate steps were taken based on gender to minimize the potential for bias in the process.
In my experience with mathematics in the U.S., these sorts of steps are pretty common. In the committees I've served on (for both admissions and hiring), people have often gone out of their way to try to identify diverse sorts of candidates and make sure they are not overlooked or disadvantaged. Not everyone participates eagerly in this, but some do it out of conviction that it's intrinsically worthwhile, while others play along to keep the administration from complaining. On the other hand, I've never seen this process extend to advantages in the final decision. In particular, I haven't seen a case in which Alice was hired or admitted instead of Bob just because she was female, although the committee was more impressed by Bob than Alice otherwise.
Tie breakers are the closest I've seen to an explicit preference. In graduate admissions many decisions are easy, but there's always a (small) group of comparable candidates right near the borderline for admission, where nobody can give a compelling argument for why one is superior to another. Within that group, being female could prove an advantage: if Alice and Bob are equally strong candidates in other ways, but Alice would help bring gender balance to the department and Bob would not, then that's a good reason to admit Alice. This doesn't generally arise in hiring, since few enough people are hired that there are many strong opinions and the hiring committee is unlikely to decide two candidates are truly tied. However, it can happen in admissions, which is a lower-stakes process carried out on a larger scale and with less information. I'm mainly mentioning it for completeness, since few applicants are actually close enough to the cut-off for this to matter.