Emission from high-J CO lines in galaxies has long been proposed as a tracer of X-ray dominated regions (XDRs) produced by AGN. Of particular interest is the question of whether the obscuring torus, which is required by AGN unification models, can be observed via high-J CO cooling lines. Here we report on the analysis of a deep Herschel-PACS observation of an extremely high J CO transition (40-39) in the Seyfert 2 galaxy NGC 1068. The line was not detected, with a derived 3$σ$ upper limit of $2 \times 10^{-17}\,\text{W}\,\text{m}^{-2}$. We apply an XDR model in order to investigate whether the upper limit constrains the properties of a molecular torus in NGC 1068. The XDR model predicts the CO Spectral Line Energy Distributions for various gas densities and illuminating X-ray fluxes. In our model, the CO(40-39) upper limit is matched by gas with densities $∼ 10^{6}-10^{7}\,\text{cm}^{-3}$, located at $1.6-5\,\text{pc}$ from the AGN, with column densities of at least $10^{25}\,\text{cm}^{-2}$. At such high column densities, however, dust absorbs most of the CO(40-39) line emission at $λ = 65.69\, μ$m. Therefore, even if NGC 1068 has a molecular torus which radiates in the CO(40-39) line, the dust can attenuate the line emission to below the PACS detection limit. The upper limit is thus consistent with the existence of a molecular torus in NGC 1068. In general, we expect that the CO(40-39) is observable in only a few AGN nuclei (if at all), because of the required high gas column density, and absorption by dust. ; Comment: 22 pages, accepted for publication in ApJ