We propose an economical model to explain the apparent 130 GeV gamma ray peak, found in the Fermi/LAT data, in terms of dark matter (DM) annihilation through a dipole moment interaction. The annihilating dark matter particles represent a subdominant component, with mass density 7--17% of the total DM density; and they only annihilate into $\ensuremath{\gamma}\ensuremath{\gamma}$, $\ensuremath{\gamma}Z$, and $ZZ$, through a magnetic (or electric) dipole moment. Annihilation into other standard model particles is suppressed, due to a DM mass splitting in the magnetic dipole case, or to $p$-wave scattering in the electric dipole case. In either case, the observed signal requires a dipole moment of strength $\ensuremath{\mu}\ensuremath{\sim}2/\mathrm{TeV}$. We argue that composite models are the preferred means of generating such a large dipole moment, and that the magnetic case is more natural than the electric one. We present a simple model involving a scalar and fermionic techniquark of a confining SU(2) gauge symmetry. We point out some generic challenges for getting such a model to work. The new physics leading to a sufficiently large dipole moment is below the TeV scale, indicating that the magnetic moment is not a valid effective operator for LHC physics, and that production of the strongly interacting constituents, followed by techni-hadronization, is a more likely signature than monophoton events. In particular, four-photon events from the decays of bound state pairs are predicted.