Every year there are more call diversion programs going live and expanding, including sending mental health calls to a local program or 988 to diverting low acuity medical calls to telehealth or a nurse call line. One challenge with these programs can be that the volume of calls being diverted is usually lower than expected, while another is the risk of true emergencies needing immediate response get sent out of the 911 system when they really need it. And all of this increases complexity for newly released telecommunicators, trainees, and veterans alike. This session will share how PSAPs are using AI models trained on specific call types/priorities to help telecommunicators send more calls while actually de-risking by decreasing the "bounce back" rate.