Updated constraints on dark matter cross section and mass are presented combining CMB power spectrum measurements from Planck, WMAP9, ACT, and SPT as well as several low-redshift datasets (BAO, HST, supernovae). For the CMB datasets, we combine WMAP9 temperature and polarization data for l <= 431 with Planck temperature data for 432 < l < 2500, ACT and SPT data for l > 2500, and Planck CMB four-point lensing measurements. We allow for redshift-dependent energy deposition from dark matter annihilation by using a `universal energy absorption curve. We also include an updated treatment of the excitation, heating, and ionization energy fractions, and provide updated deposition efficiency factors (f_eff) for 41 different dark matter models. Assuming perfect energy deposition (f_eff = 1) and a thermal cross section, dark matter masses below 26 GeV are excluded at the 2-sigma level. Assuming a more generic efficiency of f_eff = 0.2, thermal dark matter masses below 5 GeV are disfavored at the 2-sigma level. These limits are a factor of ~2 improvement over those from WMAP9 data alone. These current constraints probe, but do not exclude, dark matter as an explanation for reported anomalous indirect detection observations from AMS-02/PAMELA and the Fermi Gamma-ray Inner Galaxy data. They also probe relevant models that would explain anomalous direct detection events from CDMS, CRESST, CoGeNT, and DAMA, as originating from a generic thermal WIMP. Projected constraints from the full Planck release should improve the current limits by another factor of ~2, but will not definitely probe these signals. The proposed CMB Stage IV experiment will more decisively explore the relevant regions and improve upon the Planck constraints by another factor of ~2.