I don't think that's anywhere close to an accurate description of the UK's recent post office scandal?
I haven't seen AI feature in any reporting. Rather, the software had bugs, some people decided the software couldn't be wrong and convinced others to the point of conviction?
Indeed, that scandal vastly predated AI everywhere, and was just vanilla consultant-grade software (i.e. trash) coupled with vast incompetence at every stage.
Incompetence is not sufficient to explain that case. Many high level leaders were aware of deficiencies in the software and wanted to prosecute anyway. That is active malice or indifference to human suffering.