"The AI does not 'work out tactics'. What can sometimes change tactically is that AI teams depending on your own club's league table position and form, will look at you differently. For instance, if you're say a team like say Bournemouth in the Premier League, a large number of teams, certainly in the top 8 will look at them and consider their own team to be stronger. Because of this they're likely to play attacking formations. Mid-table sides likewise will realise they're stronger on paper and again will attack, but may take a slightly more cautious approach when they're the away team. Lower ranked Premier teams will likely play a balanced formation home and away.
Then let's say by January in your first season, Bournemouth are 5th in the table and have won impressive victories. The AI managers will look at this and realise you shouldn't be underestimated. Teams will be more likely to play a defensive counter based style when coming to play at Bournemouth and be more balanced overall.
Mid and certainly lower sides in the table will play defensively. This means a tactic you had working fine for the entire system is suddenly playing against totally different tactics than in the first half of the season. Many human managers assume that if their tactic is working there's never a need to change it, but fail to take into account the fact the AI will adapt as the season goes on. Tactics however, as said above are not individually 'worked out'.
"Why should i have to change things after so many games?... If it's not broke, why fix it?"
A manager has to realise they will need to adapt and improve their squad as time goes on. You can't just plug in a tactic, press continue and expect to win every game. The user already has a pretty large advantage over the AI when it comes to squad building (as simply the AI is never going to have as many options or be as smart as the user until we make huge technological advances in AI), but you can't expect one tactic to work against every combination of team and setup."