Let me first jump to the fact that correlation does not mean causation. Correlation can be an indicator, but correlation itself does not prove a relationship. That would be like claiming linear contemporaneous forward travel in time at the normally accepted rate causes a population explosion in humans.
Over that last 15+ years I have noticed a reduction in the amount and proliferation of a type of reed on the river we locally just call pencil tulies.
Some time back 25+ (noticed in 1999) years ago we started having a problem with a with an invasive species call giant salvinia. A weevil was introduced (2001) to combat giant salvinia, and it appears to be doing a good job.
I generally started noticing the reduction in pencil tulies around 2010. I'm not sure when it started, but I thought it might be a year to year variation, but every year since there has been less and less.
I asked Google search if the weevil was having an affect on any vegetation other than salvinia, and it was an emphatic, no.
After asking many questions each one framed to contain the previous question and ask for more specific results. The AI's last (useful) response was speculation that it was caused by reduced water levels and poor water quality due to human causes. (generally) The water quality has generally not changed in the lower Colorado River. In the lower river the water level fluctuates constantly due to farm demand. It is not controlled (except during massive floods) by available water, some areas fluctuate very little, and due to very controlled management one lake doesn't fluctuate more than an inch or two, and then only during extreme weather or during the rare repairs of facilities.
Now I realize there may be a number of political issues wrapped up in both my questions and the subject matter at large, but my questions were very carefully nonpolitical asking only about cause and affect in the ecosystem.
More answers from AI just stomped its feet and said salvinia weevils didn't affect native species. Then I caught the nuance. Maybe the reeds or pencil tulies are not a native species.
I decided to see if I could determine if it was a native plant. I tried to determine if the AI could give me an answer, and simply trying to get an answer is difficult. It kept giving me wrong or misleading answers.
While I have no definitive proof, I am lead to be believe that when the initial responses wanted proof and clarification the AI was deliberately evading the question.
Now two things. My original thought that the salvinia weevil might be attacking other plants as its primary food source is gobbled up, and that one followed the other is not in itself any sort of proof. Just a hypothesis, and correlation does not mean causation.
Second is that I have a healthy anti AI bias. I do use results, but I try to always look at the sources. This I believe is a healthy bias based on direct observation. Not just prejudice.