In addition, If FFH (and other insurers) have not explicitly excluded certain risks (or implicitly identified the risks they will cover to the exclusion of everything else), risks like injury due to a AI bias (e.g. a FFH client is sued by their customer/consumer claiming injury due to the FFH client implementing AI) or inability to attribute fault (e.g. AI controlled system crashes. Who among the owner, operator or manufacturer is at fault?) may add up. Also, there are lawyers out there who specialize only in "mass claims". AI might make their lives easier by discovering plaintiffs faster.