The muon g μ − 2 discrepancy between theory and experiment may be explained by a light vector boson Z d that couples to the electromagnetic current via kinetic mixing with the photon. We illustrate how the existing electron g e − 2, pion Dalitz decay, and other direct production data disfavor that explanation if the Z d mainly decays into e þ e − , μ þ μ − . Implications of a dominant invisible Z d decay channel, such as light dark matter, along with the resulting strong bounds from the rare K → π + missing energy decay are examined. The K decay constraints may be relaxed if destructive interference effects due to Z − Z d mass mixing are included. In that scenario, we show that accommodating the g μ − 2 data through relaxation of K decay constraints leads to interesting signals for dark parity violation. As an illustration, we examine the alteration of the weak mixing angle running at low Q 2 , which can be potentially observable in polarized electron scattering or atomic physics experiments.