Recently, differences in Doppler shifts across the base of four close classical T Tauri star jets have been detected with the HST in optical and near-ultraviolet (NUV) emission lines, and these Doppler shifts were interpreted as rotation signatures under the assumption of steady state flow. To support this interpretation, it is necessary that the underlying disks rotate in the same sense. Agreement between disk rotation and jet rotation determined from optical lines has been verified in two cases and rejected in one case. Meanwhile, the NUV lines, which may trace faster and more collimated inner spines of the jet than optical lines, either agree or show no clear indication. We propose to perform this test on the fourth system, Th 28. We present ALMA high angular resolution Band 7 continuum, 12CO(3-2) and 13CO(2-1) observations of the circumstellar disk around the T Tauri star Th 28. We were able to detect, in CO and continuum, clear signatures of a disk in Keplerian rotation around Th28. The 12CO emission is resolved, allowing us to derive estimates of disk position angle and inclination. The large velocity separation of the peaks in 12CO, combined with the resolved extent of the emission, indicate a central stellar mass in the range 1-2 Msun. The rotation sense of the disk is well detected in both 13CO and 12CO emission lines, and this direction is opposite to that implied by the transverse Doppler shifts measured in the optical lines of the jet. The Th 28 system is the second system where counter-rotation between the disk and the optical jet is detected. These findings imply either that optical transverse velocity gradients detected with HST do not trace jet rotation or that modeling the flow with the steady assumption is not valid. In both cases jet rotation studies that rely solely on optical lines are not suitable to derive the launching radius of the jet.