Hello, I've recently discovered that AI seems to be quite good at identifying instances of manipulative language via text (text messages, emails, etc.) and I was wondering if someone with a more techy background could tell me truly how effective AI is at this and whether it would be admissible in court for a divorce/child custody case? I imagine how the data generated from AI is presented plays a big part in admissibility. I was thinking of having AI comb through about 7 years worth of text messages and emails, flagging and analyzing any potentially manipulative or abusive language and then bring those flagged conversations to a 3rd party therapist/psychologist for them to do an unbiased (as unbiased as possible) analysis of the AI flagged text to corroborate with a real trained person. For context, I'm divorcing a narcissist who has become increasingly talented at manipulation, isolation, and turning those close to me against me. Ideally I would have witnesses but apparently my "friends" believe her lies and half truths more than they believe my cries for help. Any info/guidance/help is appreciated.