Organisation Name

New York University Shanghai

Industry

Public service

Organisation Website

https://shanghai.nyu.edu/

Country

China

Sustainable Development Goals (SDGs)

General Description of the AI tool

AI Supported Questioning (ASQ) is a fine-tuned large language model that provides real-time, context-sensitive question suggestions during child investigative interviews. Trained on 649 high-quality transcripts following the NICHD protocol, ASQ outperforms both trained interviewers and base models by generating significantly more open-ended and non-suggestive questions while minimizing yes–no formats. It reduces interviewer cognitive load, enhances accuracy and completeness of children’s testimony, and ensures global access to evidence-based child protection practices.

Relevant Research and Publications

1. Haginoya, S., Sun, Y., & Santtila, P. (2025). Retention of training effects in police officers following scalable AI simulations of child sexual abuse interviews. Behavioral Sciences & The Law.
2. Kask, K., Pompedda, F., Palu, A., Schiff, K., Mägi, M.-L., & Santtila, P. (2022). Transfer of avatar training effects to investigative field interviews of children conducted by police officers. Frontiers in Psychology, 13, 753111.
3. Pompedda, F., Zhang, Y., Haginoya, S., & Santtila, P. (2022). A mega-analysis of the effects of feedback on the quality of simulated child sexual abuse interviews with avatars. Journal of Police and Criminal Psychology, 37, 485–498.
4. Haginoya, S., Yamamoto, S., & Santtila, P. (2021). The combination of feedback and modeling in online simulation training of child sexual abuse interviews improves interview quality in clinical psychologists. Child Abuse & Neglect, 115, 105013.
5. Santtila, P., Sun, Y., Järvilehto, L., Antfolk, J., Haginoya, S., Pakkanen, T., Lamb, M. E., & Korkman, J. (2025). From training towards real-time support: Fine-tuned AI steers question suggestions toward invitations and wh- questions in a Child Sexual Abuse interviewing task. Manuscript in preparation.

Needs