The SPI spectrometer aboard of the INTEGRAL satellite has released a map of the e+e− annihilation emission line of unprecedented quality, showing that most of the photons arise from a region coinciding with the stellar bulge of the Milky Way. The impressive intensity (≃ 10-3 photon cm-2 s-1) and morphology (round and wide) of the emission is begging an explanation. Different classes of astrophysical objects could inject positrons in the interstellar medium of the bulge, but the only acceptable ones should inject them at energies low enough to avoid excessive bremsstrahlung emission in the soft gamma ray regime. Among the ~ MeV injectors, none seems generous enough to sustain the high level of annihilation observed. Even the most profuse candidate, namely the β+ radioactivity of 56Co nuclei created and expelled in the interstellar medium by explosive nucleosynthesis of type Ia supernovae, falls short explaining the phenomenon due to the small fraction of positrons leaking out from the ejecta (≈3%), together with the low SNIa rate in the bulge (≈0.03 per century). It is therefore worth exploring alternative solutions, as for instance, the idea that the source of the positrons is the annihilation of light dark matter (LDM) particles of the kind recently proposed, totally independently, by Bœhm and Fayet. Assuming that LDM is the culprit, crucial constraints on the characteristics (mass and annihilation cross-section) of the associated particle may be discussed, combining direct gamma ray observations and models of the early Universe. In particular, the mass of the LDM particles should be significantly less than 100 MeV, so that the e+ and e− resulting from their annihilations do not radiate exceedingly through bremsstrahlung in the interstellar gas of the galactic bulge.