Abstract Dust close (∼few pc) to the accretion disk in active galactic nuclei (AGNs) is key to understanding many of their observational signatures, and it is key to tracing how the AGN is fed or even evolves along its duty cycle. With estimated sizes of less than 10 pc, as constrained by mid-IR (MIR) high angular resolution data, only the superb spatial resolution achieved by ALMA is able to actually image this dusty structure. However, the question regarding how the dust at submillimeter wavelengths (sub-mm, typical ALMA band) behaves in the AGN contest, arises. We study the detectability of the emission associated with the AGN dusty structure at submillimeter wavelengths using ALMA, in a theoretical and observational way. Theoretically, we use the Clumpy models from Nenkova et al., together with the MIR to X-ray and radio fundamental plane scaling relations. We find that it is more likely to detect bigger and denser dusty tori at the highest ALMA frequency (666 GHz/450 μm). We also find that with 1 hr at 353 GHz/850 μm and 10 hr at 666 GHz/450 μm we can detect, with a high detection limit, a 1 mJy torus (characteristic of bright AGNs). This means that an object for which the unresolved spectral energy distribution (SED) is at 12 μm has a flux of ∼1 mJy. Observationally, we use four prototypical AGNs: NGC 1052 (low-luminosity AGN), NGC 1068 (type 2), NGC 3516 (type 1.5), and I Zw 1 (QSO), with radio, submillimeter, and MIR data available. All the MIR spectra are best fit with the smooth model reported by Fritz et al. A power law and a single, or a composition of, synchrotron component(s) reproduce the centimeter radio wavelengths. We combined and extrapolated both fits to compare the extrapolation of both torus and jet contributors at submillimeter wavelengths with data at these wavelengths. Our observational results are consistent with our theoretical results. The most promising candidate to detect the torus is the QSO I Zw 1 (therefore, highly accreting sources in general), although it cannot be resolved owing to the distance of this source. We suggest that to explore the detection of a torus at submillimeter wavelengths, it is necessary to perform an SED analysis including radio data, with particular attention to the angular resolution.