RL, including contextual bandits, is becoming more popular for personalization, i.e. adapting some system to the preferences of (groups of) individuals.

Plug/Source: I did a lit. review on this topic https://doi.org/10.3233/DS-200028