[en] If advanced extraterrestrial civilizations choose to construct vast numbers of Dyson spheres to harvest radiation energy, this could affect the characteristics of their host galaxies. Potential signatures of such astroengineering projects include reduced optical luminosity, boosted infrared luminosity, and morphological anomalies. Here, we apply a technique pioneered by Annis to search for Kardashev type III civilizations in disk galaxies, based on the predicted offset of these galaxies from the optical Tully–Fisher (TF) relation. By analyzing a sample of 1359 disk galaxies, we are able to set a conservative upper limit of $\lesssim <!--\; ???\; -->3\mathrm{\%<!--\; \%\; -->}$ on the fraction of local disks subject to Dysonian astroengineering on galaxy-wide scales. However, the available data suggests that a small subset of disk galaxies actually may be underluminous with respect to the TF relation in the way expected for Kardashev type III objects. Based on the optical morphologies and infrared-to-optical luminosity ratios of such galaxies in our sample, we conclude that none of them stand out as strong Kardashev type III candidates and that their inferred properties likely have mundane explanations. This allows us to set a tentative upper limit at $\lesssim <!--\; ???\; -->0.3\mathrm{\%<!--\; \%\; -->}$ on the fraction of Karashev type III disk galaxies in the local universe.