The author of the article has named the phenomenon incorrectly. Adsorption can indeed result in a lower volume than might be expected, for example when a gas interacts with a high surface area solid. In the case of hydrogen storage, for example, the adsorbed phase is dense enough at low temperatures that much more hydrogen can be stored in a container filled with a high surface area material, and at lower pressures, than an empty container of the same volume.
However, the situation that is described actually is a result of solvation, not adsorption, and the volume-reducing effect has been discussed in some. other. questions, have a look around!
In essence, adsorption is only discussed in the context of the interface between two phases, such as gas-liquid, gas-solid, or two immiscible liquids.
If someone were feeling really pedantic, they could make an argument that the solvated molecules in the liquid constitute a really small "adsorbent" particle, and the solvent molecules in the first solvation shell are "adsorbates", and more power to them. Really, the forces that are at work in both cases (solvation and adsorption) are very similar.
It might be an interesting study to investigate where the crossover between adsorption and solvation happens as a function of molecule size. After all, graphene is just a really big molecule and its adsorption properties are pretty great. A brief search on adsorption solvation crossover
reveals that there is some investigation that has been done on the process of going from a 3D gas $\ce{->}$ 2D adsorbate $\ce{->}$ aqueous solvate here.
tl;dr adsorption does occur at the interface between immiscible liquids, but the OP's article is talking about solvation