Latent Profiles of Technology Facilitated Sexual Harm of Minors

Background/Purpose: Technology-facilitated sexual harm (TFSH) of minors is an ongoing public health crisis. The emergence of interactive social media platforms and, more recently, generative AI, have created efficient avenues for offenders to access minors and new tools for the creation and distribution of child sexual abuse material (CSAM). Governments across the globe have reported large increases in AI-generated CSAM and struggled to implement policies that effectively protect children online. Given the emergence of new technologies used for abuse and a need to develop strategies for preventing online violence, this study aimed to identify patterns of TFSH among minors. This study also examined associations between patterns of TFSH and social and interpersonal risk factors, the identities of the person(s) who offended TFSH against minors, and demographic characteristics of minors who experience harm.

Methods: Analyzed data were collected as part of a nationwide study of young people’s positive and negative experiences online (n = 1,874). The present analytic sample consisted of individuals who experienced TFSH online before the age of 18 (n = 451). MPlus was used to conduct latent class analysis with eight forms of TFSH (e.g., deepfake creation, sharing of sexual images or videos, offers of money or gifts for sexual actions, etc.) as latent indicators. After selecting the models that best fit the data, the 3-step BCH process was used to conduct latent class multinomial regressions to determine how social and interpersonal risk factors, offender identities, and demographic characteristics of minors were associated with latent class membership.

Results: A four-class solution best fit the data. Classes were labeled as “sextortion” (46%; experienced non-consensual recording of sexual content and/or threats to share sexual content), “solicitation” (22%; experienced being offered money or gifts for sexual images, videos, or actions), “polyvictimization” (18%; experienced multiple forms of victimization), and “synthetic media” (14%; experienced non-consensual sexual image generation).

Using “sextortion” as the reference class, social and interpersonal risk factors, offender identities, and demographic characteristics were associated with class membership. Compared to the “sextortion” class, “solicitation” class membership was associated with having fewer online sexual experiences, “polyvictimization” class membership was associated with experiencing more adverse childhood experiences and lower socioeconomic status, and “synthetic media” class membership was associated with fewer protective childhood experiences.

Compared with “sextortion,” offenders for those in the “solicitation” class were more likely to be strangers, “polyvictimization” class offenders were more likely to be strangers or online connections (with or without connections to peers), and family members were more likely to be “synthetic media” offenders. Romantic partners/exes were more likely the offenders of sextortion compared to synthetic media or solicitation.

Conclusions/Implications: Findings suggest patterns exist in minors’ experiences of TFSH, and these patterns may be predicted by social and interpersonal risk factors as well as offender identity. The existence of unique patterns and risk factors in TFSH has implications for the development of tailored prevention programs that intervene upon social and interpersonal risk factors as well as for practitioner responses to support individuals who have experienced multiple forms of TFSH.