I don't believe Ag can react with TiO2. Both are already in their favoured state. In the quoted paper, something else must decompose AgNO3, and I bet for ethanol.
The reagents in that paper resemble a bit Tollens' ones, where the reaction is not Ag+ receiving an electron. It first builds a complex.
As the quoted paper deposits Ag on pure TiO2, you shouldn't worry about Ti's oxide layer. Which to my eyes speaks in favour of a catalytic deposition, not a redox.
Thin enough native oxide on Ti, or elsewhere, does not prevent current flow. A good oxide layer (ultra-pure Al, Ta, Ti, Nb...) has a breakdown voltage, below which it insulates, and above it conducts (the curve is fuzzy of course). The native layer has a faint breakdown voltage, so if the (ultra-pure) electrolyte builds a clean oxide layer, the layer grows until the resulting breakdown voltage equals the applied one.
Consequently, the layer is very uniform, and its thickness just fits the forming voltage. This is an excellent way to produce capacitors for low-voltage and high capacity, where plastic films would be too thick. Al, Ta, Nb are used for that purpose.
It's the reason why I suggested electrodeposition. Though, you chances of success depend on how the oxide film conducts electricity: by electrons, or by transport of oxygen ions, or transport of titanium ions, or transport of protons...
With silicon we use (used, I stopped long ago) HF (very hazardous indeed, better keep off) to remove oxide layers, but the native layer reforms as soon as (<<1s) silicon isn't wet any more; or even, it forms continuously in the liquid and gets etched as quickly.