(064) Algorithmic Sexual Health: Use of AI for Prevention of Sexual Abuse and Professional Training

Introduction Artificial intelligence (AI) has emerged as a potential valuable addition to the practice of sexual medicine and in public health, both in terms of clinical practice, research methods, and training approaches. While the value is presumed to be substantial, little is known about its efficacy in the actual real-world application. Objective The presentation aims to present examples of AI use in the field of sexual health. The first will be the AI-assisted online intervention aimed at users of child sexual abuse material (CSAM), discussing the key results and potential future applications of this technology. The second will be a new project, KOMPAS, aimed at the training of professionals in the field of sexual health via AI-powered digital avatars. These case studies will illustrate the potential of AI in the field. Methods The AI-assisted intervention ("STOP CSAM") was evaluated using a waitlist-control design, following a group of participants over a 16-week period, focusing on the reduction in frequency and severity of the consumed CSAM. Standard techniques used in RCTs have been used. The KOMPAS project's avatars will be evaluated by using a control group of students and professionals receiving standard training, and comparing their skills and feedback to those exposed to the intervention. Usability will be tested using standard scales such as the System Usability Scale. Results The results of the STOP CSAM AI-assisted intervention point to promising results, with the participants showing reduced usage of CSAM following the treatment compared to the control group. The therapists have used the AI to guide the interactions with the clients, helpling them generate alternative responses and summarize past interactions. The added value of the KOMPAS avatars is not known yet, but initial results from professional focus groups will be shared at the conference. Conclusions The AI assistant seems to be a promising addition to similar interventions delivered via online chat, although its effectiveness should be tested in every future therapeutic setting. In general, using new AI tools in the field of sexual health and prevention of sexual offending offers a promising path to a greater efficiency of the system, although attention must be paid to ensuring its use is responsible, transparent, and rigorously evaluated. AI tools will not be immediately revolutionary in the field, but they have the potential of solving many partial problems and moving sexual medicine forward. Disclosure No