Of course they don’t. 
I’m not a game developer, but as I understand it, Paradox games are at the deep end of the gaming industry in terms of complexity. They are more like simulations than “games.” To top it off, the major games, with the exception of Stellaris and CS2, are based on history, which makes the use of generic AI in content… more complicated. Games are also developed for years before release. Yet, they are practically in beta at launch before years of DLCs and fixes. 
Paradox’s DLC model allows for refining games to their peak over the years in a financially sensible way.
That’s why I wouldn’t be primarily worried about AI disruption in games regarding Paradox, unless the entire player base shifts to lighter games. For example, @Farseer raised this concern well based on their own experiences. It’s possible that the learning curve for Paradox games is becoming too high for many. On the other hand, that makes the games more rewarding when they offer challenges and insights…
And shouldn’t the same AI concern perhaps be visible first in the giant of the strategy space, i.e., the Civilization game series? That’s where there is the most meat for challengers to go after.