By Margaret (Maggie) Brennan Lecturer at University College Cork, Ireland
There is a dearth of comprehensive and consistent data on the characteristics of Child Sexual Abuse Material (CSAM), Child Sexual Exploitation Material (CSEM), and the children and offenders depicted in this content. This situation is due, in part, to the criminal nature of these materials, related methodological, ethical and legal challenges in researching CSAM/CSEM, and highly limited resourcing in relation to the significance of the issue. Until recently, no representative international baselines of empirical data had been produced on these phenomena.
In an effort to address this problem, I was commissioned by ECPAT International and INTERPOL to lead an analysis of a multi-country data set of CSAM/CSEM cases, seized by law enforcement around the world, and housed in the International Child Sexual Exploitation (ICSE) Database at INTERPOL. This CSAM/CSEM data set is broader in country coverage than any other previously analysed and made public. Funded by the European Commission, the study presents the results of a two-part analysis of CSAM/CSEM case data in the ICSE Database, and of consultations with law enforcement involved in the identification of victims and offenders pictured in CSAM/CSEM. It highlights the multi-faceted challenges presented to international law enforcement, child protection and other stakeholders by rapid evolutions in online child exploitation and abuse, and the increasingly complex role played by’ youth-produced’ sexual content in this context.
A major element of the report was the analysis of CSAM/CSEM images and videos, undertaken to develop a descriptive profile of unidentified child victims and their abusers, with attention to variables such as age category, gender, ethnicity, type and severity of depicted sexual activity, and paraphilic theme. This involved a visual analysis of a random sample of 800 CSAM/CSEM series drawn from unidentified cases in the ICSE Database. Data collection from the CSAM/CSEM series was guided by a bespoke, 22-category coding framework, which in turn, was subjected to an interrater reliability assessment.
Evidently, analysing recordings of child sexual victimisation raised many complex ethical challenges for the research partners, particularly from the perspective of child rights. Therefore, the study was subject to a wide range of legal, institutional and ethical conditions, duly and carefully considered, and rigorously implemented, in order to respond to the ethical issues the project raised. These are described in the full report.
Very young children
The analysis suggested that, in comparative terms, the situation of very young CSAM/CSEM victims was particularly acute. The relationship between the severity of depicted sexual activity and victim age was significant, with infants and toddlers more likely to feature in imagery depicting severe sexual abuse involving an adult (COPINE level 8-10). Furthermore, very young children were more likely than children of other ages to be subjected to victimisation featuring an additional problematic paraphilic theme (i.e. additional to the more obvious paedohebephilic themes depicted in the imagery). Overall, these additional problematic paraphilias were depicted in almost one third of the analysed CSAM/CSEM series.
Boys as victims
31% of series depicted the victimisation of boys exclusively. This figure is substantially higher than those reported in other studies, where boy victims accounted for approximately 20% of analysed cases (e.g. Canadian Center for Child Protection on the Internet, 2016; Quayle & Jones, 2011). Moreover, there was a significant relationship between the severity of depicted sexual activity and victim gender, with boys more likely to feature in material depicting severe sexual abuse (COPINE level 7-10), and girls more likely in imagery depicting moderate victimisation (COPINE level 4-6).
Children in ‘low-level’, sexualized imagery
Over 61% of analysed series were identified as being both ‘abusive’ and ‘exploitative in character’, meaning that universally illegal sexual abuse images and potentially legal exploitation images of the same victim were found together. This finding speaks to the possibility that many child subjects of ‘low level’ exploitation imagery have also been implicated in the production of illegal CSAM.
Where depicted, females offended more frequently alongside a male, assuming an ‘active’ role in the abuse. Moreover, offending pairs comprising male and female offenders were more likely to engage in extreme forms of abuse. There was significant relationship between offender gender and sexual activity level, with series where males and females depicted together more likely to feature the highest level of abuse (COPINE level 10).
‘Youth-produced’ sexual imagery
A wide range of sexual activities were depicted in these materials, from more innocuous, nude or semi-nude ‘selfies’ of children, through to ‘self-generated’ depictions of extreme sexual activity involving bestiality and sadomasochistic themes. While many recordings were produced in domestic settings, others were apparently produced in school environments. The levels of CSAM/CSEM production depicted in these cases were quite complex, and challenged the simplistic distinction that has been drawn between content that is ‘youth-produced’ and offender-generated. In some cases, offender involvement was clear, whether recording the children while they ‘self-generated’ the imagery, or otherwise coercing the child into the production of the content.
Given the scope of the analysis, the findings presented here are limited to a high-level selection of report highlights which may be of interest to those concerned with the treatment and management of online child sex offenders and their victims.
More broadly, the study highlighted how our knowledge of the characteristics of CSAM/CSEM victims and offenders is limited, both by a lack of standardised or comparable data categorisation approaches, and by differences in the sampling and case recording approaches across existing studies. Resolving this situation will require extensive engagement between the research community and gatekeepers of international repositories of CSAM/CSEM in order to develop standardised and comparable datasets. Notwithstanding, the study offers a framework and categorisation approach towards this goal, that may be further used to support the development of descriptive profiles of CSAM/CSEM victims and offenders in future studies.
The technical report, containing full findings and discussion, can be downloaded from: http://www.ecpat.org/wp-content/uploads/2018/02/Technical-Report-TOWARDS-A-GLOBAL-INDICATOR-ON-UNIDENTIFIED-VICTIMS-IN-CHILD-SEXUAL-EXPLOITATION-MATERIAL.pdf
Margaret (Maggie) Brennan Lecturer at University College Cork, Ireland; research lead on the ECPAT-INTERPOL study, ‘Towards a Global Indicator on Unidentified Victims in Child Sexual Exploitation Material’.
Canadian Center for Child Protection (2016). Child Sexual Abuse Images on the Internet: A Cybertip.ca Analysis. Retrieved from https://www.cybertip.ca/pdfs/CTIP_CSAResearchReport_2016_en.pdf
Quayle, E., & Jones, T. (2011). “Sexualised Images of Children on the Internet”. Sexual Abuse: A Journal of Research and Treatment, 23(1), 7-21.
Taylor, M., Holland, G., & Quayle, E. (2001). “Typology of paedophile picture collections”. The Police Journal, 74(2), 97-107.
 The Interagency Working Group on Sexual Exploitation of Children (2016) defines CSAM as a subset of CSEM, ‘where there is actual abuse or a concentration on the anal or genital region of the child’ (p.39). According to its guidelines, CSEM can be used in a broader sense to encompass all other sexualised material depicting children (p. 40).
 A series is a group of images and/or videos that are related to each other in some way that is meaningful to an investigator, e.g. if the images and videos depict the same victim or the same crime scene.
 Levels of agreement between the raters for the framework categories (reliabilities) were measured by means of an assessment of inter-rater reliability using Kendall’s tau (τ). Where it was possible to produce reliability estimates, scores indicative of good to perfect agreement were observed between the 4 raters in the application of the framework.
 Severity of depicted sexual victimisation was assessed in accordance with the 10-point COPINE scale (Taylor, Holland & Quayle, 2001)