How intelligent is the AI about analyzing the activity? I've worked with some of those tools before, and if this AI was onto them, it'd make things miserable. The QA/software test team would be submitting poor quality issues and bugs just for the sake of "activity". I guess that would help the developers too because they can fix these issues and get a better rating according to this "AI".
https://dilbert.com/strip/1995-11-13
Reminds me of when IBM paid by the "kloc" (thousands of line of code). It would just encourage bad, inefficient code, for the sake of higher pay. I believe the US government did the same thing with the railroad companies a couple centuries back.. they paid by the mile, not taking into proper account for efficiencies.