Aharonov and Bohm noted that the wave-function of a charge acquires a detectable phase when superposed along two paths enclosing an infinite solenoid, even though that wave-function is nonzero exclusively where the solenoid's electric and magnetic fields are zero. This phase was long considered radically different from all other quantum phases, because it seems explainable only via local action of gauge-dependent potentials, not of the gauge-invariant electromagnetic fields.Recently Vaidman proposed a model explaining the phase only in terms of the electron's (magnetic) field at the solenoid. Later Kang proposed a Lagrangian model from which Vaidman's can be derived. However, their analysis is still non-local. For it does not explain how the phase, generated at the solenoid, is detectable on the charge when closing the interference loop. To explain this, one needs to quantise the EM field and consider how the interaction between the solenoid and the electron is mediated by photons. In this paper we propose a local model, where the field is quantised. This shows that the Aharonov-Bohm (AB) phase is generated by the same quantum local mechanism as all other electromagnetic phases. Specifically, it is mediated by the entanglement between the superposed charge and the photons. Surprisingly, the quantised model produces some experimentally different predictions from the semiclassical accounts of the AB phase, because it predicts that the phase at any point along the charge path is gauge-invariant and locally generated, and therefore in principle detectable. We propose a realistic experiment, within current technological reach, where the predicted phase difference along the path is measured locally, by performing (a partial) quantum state tomography on the charge without closing the interfering charge paths.