Vol. XXVIII, No. 3
Summer 2016
Text Only Version
In This Issue
Regular Features
Editor's Note
President's Message
Are Juvenile Sexual Risk Assessment Instruments Adequate on Their Own to Assess Risk?
Featured Articles
Preventing Clinician Burnout
A Theoretical Framework for Proscribing Pornography Viewing for Those With Sex Offense Convictions
Online Debate 5: Developing a worldly understanding of sexual offenders and their management
Students' Voice
Sexual Deviance and General Criminality Factors Among Adolescent Sex Offenders
3rd Annual ATSA Student Clinical Case and Data Blitz
Book Review
The Trauma Myth
2016 Election
35th ATSA Conference
Awards Announcements
ATSA Chapters: Amplifying ATSAís Footprint in the World
New ATSA Members
Newsletter Tools
Search Past Issues
Print-Friendly Issue
Print-Friendly Article
Forum Team
David Prescott
Book Review Editor

Sarah Gorter
Production Editor

Forum Editor
Contact the editor or submit articles to:

Heather M. Moulden, Ph.D.
Forensic Program
St. Joseph's Healthcare
Hamilton, Ontario, Canada
E: hmoulden@stjoes.ca
P: (905) 522-1155 ext. 35539
Online Debate 5: Developing a worldly understanding of sexual offenders and their management
By Kieran McCartan

What follows is the transcript of the fifth online debate, titled “Developing a worldly understanding of sexual offenders and their management” which occurred as part of The Leverhulme Trust funded Community Engagement and Partnership Working with Sexual Offenders International Network. This network is a collaboration of seven academics from six universities across five countries. We also have a number of national and international partners. It is an international partnership, designed to engage practitioners, academics and policy makers on issues of sexual offender risk management and public protection. The network is also committed to public and societal education on issues of sexual offender etiology, offending, management and reintegration.

This represents the fifth of six discussions scheduled to take place throughout 2014 - 16 to examine community engagement and partnership working with sexual offenders. The debates are a series of discussions between invited practitioners, academics and relevant parties on a variety of issues relating to sexual offender management and reintegration. All of the debates will be published in the ATSA Forum as well as on the international networks website:


The aim of the current debate is to discuss streamlining and articulating an accessible public health approach to sexual violence. The debate included below occurred over a two -week period in April 2016, through an online social networking site.

Twelve participants from 3 countries (UK, USA, and Canada) agreed to participate in the debate. They were invited to take part because of their interest and expertise in this area. The participants were selected from academic backgrounds and practitioner groups. All participants are identified below only via the institutions or organisations on whose behalf they spoke:

- Participant 1: A representative from Internet Watch Foundation, UK. IWF

- Participant 2: Professor Criminology and Criminal Justice, DMU, UK. DMU

- Participant 3: Professor in the School of Criminology and Justice Studies at the University of                 Massachusetts, Lowell, USA. UoML

- Participant 4: Assistant Professor in Psychiatry, ‎McMaster University, Canada. McM 1

- Participant 5:  Adjunct Professor in Psychiatry, McMaster University, Canada. McM 2

- Participant 6: Postdoctoral Fellow, Royal’s Institute of Mental Health Research, University of      Ottawa, Canada. UoO

- Participant 7: Research psychologist, School of Applied Psychology, University College Cork,        UK. UCC

- Participant 8: Associate Professor in Psychiatry, University of Toronto, Canada. UoO

- Participant 9: Associate Professor in Criminology, UWE, UK. UWE

- Participant 10: Assistant Professor in Psychiatry, ‎McMaster University, Canada. McM 3

- Participant 11: Senior Lecturer in Law, University of Hull, UK. UoH

- Participant 12: Senior Lecturer in Psychology, University of Lincoln, UK. UoL

The debate below is a faithful representation of what was said, nothing has been otherwise altered (except in a very few places where minor adjustments were made for clarity and the removal of identifying features/comments —Editor). The format includes an opening statement, in this case by the representative from IWF, followed by an open discussion including all participants, and the debate is then summarised. Readers are reminded that this was originally an online discussion, and that grammar, etc. were of lesser importance in favour of free expression of thought and opinion.


Dear Participants,

Welcome to the online debate on developing a ‘worldly’ understanding of sexual offenders and their management.

It goes without saying that this topic covers a wide range of areas in terms of the actual offences, the management of offenders or the potential prevention efforts.

Previous online debates already provided valuable information on various aspects of offender management and I hope this session will contribute to the broader discussion by identifying examples of recent developments in sexual offending as well as potential new approaches or initiatives for offender management.

In order to keep this discussion focussed - and also to address an issue that perhaps hasn’t been addressed sufficiently in the other online debates - I would like to suggest to mainly focus on the increasing importance of cyberspace in sexual offending, in particular in relation to online child sexual abuse and online child sexual exploitation.

I will start the debate by introducing the work of the Internet Watch Foundation, the UK hotline for reporting online child sexual abuse imagery and by raising a couple of questions in relation to online offender management. You are of course free to provide your answers to these questions, to share your expertise in other areas or to address the wider topic of this online debate.

As regards the work of the Internet Watch Foundation, their team of analysts encountered 31,266 webpages depicting child sexual abuse imagery in 2014. These webpages typically include images of children under the age of 10 (80% of the webpages) and often depict the highest level of abuse, including rape and sexual torture (43% of the webpages) (www.iwf.org.uk).

The IWF works closely with UK law enforcement and the online industry to disrupt the availability of indecent images and videos of children, and is constantly working to develop new tools to find and remove these images as quickly as possible. It works with partners globally to achieve this, as only 0.3% of the webpages the IWF encounters are actually hosted in the United Kingdom.

Besides the physical abuse of the child, it is an offence to take, permit to be taken, make, possess, show, distribute or advertise indecent images of children. The Child Exploitation and Online Protection Command mentioned in their threat assessment of 2012 that in the United Kingdom around 50,000 individuals were downloading and sharing indecent images of children. (CEOP Threat Assessment of Child Sexual Exploitation and Abuse – June 2013).

The data above shows that the online dimension of sexual offending provides us with a specific set of challenges in relation to sexual offences and offender management.

-          How do we deal with the volume of people accessing online child sexual abuse material? Can we (should we?) prioritise high risk offenders within the wider group of offenders? Are existing threat assessments adequate? Is there enough information/research regarding the potential transition from online to offline abuse?

-          What should the response be towards the offenders? Is it possible to arrest all offenders or should alternative initiatives be explored? And if so, what could these initiatives look like?

-          What is the role of prevention and support for potential offenders? Organisations like ‘The Lucy Faithfull Foundation’ in the United Kingdom provide access to appropriate sources of support and intervention to help potential offenders manage their thoughts and behaviour and to assist them to live responsible, fulfilling and law-abiding lives.

-          Is the legal framework adequate and up-to-date to deal with the various aspects of the problem, for instance in relation to young people (children) sending or possessing sexually explicit images of themselves or their partner?

The introduction above provides only a snapshot of some of the issues that could be discussed in our online debate.  I would like to invite you to share your views on the overall issue of online sexual abuse and offender management as well as to share more detailed information about areas in which you might be involved or have additional expertise.


Thank you IWF for a very important introductory posting to this critical issue.  I will have an initial go at addressing some of the questions below:

- How do we deal with the volume of people accessing online child sexual abuse material? Can we (should we?) prioritise high risk offenders within the wider group of offenders? Are existing threat assessments adequate? Is there enough information/research regarding the potential transition from online to offline abuse?

Certainly UK police and probation services are concerned about how to address the sheer volume of internet offending, but also with how to identify those most likely to go on to commit contact offences.  I think there are some concerns/cautions about this and some positives.  The concerns/cautions first:

1. The distinction between internet and contact offending can be somewhat artificial- a contact crime has occurred to enable filming, some of it gross and terribly harmful.  Can we within this identify high risk offenders?  To some extent this can be done by targeting those who have used the most serious images, who have the greatest volume and frequency of use.  But in a sense this relies on existing knowledge of their pattern of use, and can work for those about whom we have existing knowledge of their offending pattern.

2.Can we identify those who will use internet offending to lead into contact offending within their own lives?  There is some potential here and an emerging evidence base.  For example, the “Fighting International Internet Paedophilia (FIIP) Project” which is led by Prof Laurence Alison at Liverpool University.  They have produced in collaboration with Kent Police the “Kent Internet Risk Assessment Tool (KIRAT) (Version 2)” which is being piloted by a number of English police forces.  Based on early evidence it appears to have a good predictive rate about those who do go onto commit contact sexual offences.  The Tool assists in prioritizing offenders for interventions and guides the management levels applied to them by police, probation and MAPPA (Multi Agency Public Protection Arrangements).  Further information can be obtained from the Child Exploitation Investigation Team and Kent Police and on kirat@kent.pnn.police.uk or:



There is a strong commitment to evaluation of the tool and revision over time, and it appears to provide a very good starting point in this challenging area.  Do people have experience of other tools or other research in this area?

- What should the response be towards the offenders? Is it possible to arrest all offenders or should alternative initiatives be explored? And if so, what could these initiatives look like?

We cannot assume that the criminal justice route is the only option for this number of offenders, particularly at the lower risk end.  What else can we do?  Projects such as Dunkenfeld in Germany offer an alternative, encouraging offenders and potential offenders to seek treatment.  Lucy Faithfull was mentioned by IWF, and Helplines are important, but so are the treatment and interventions to back it up.  At present we know that many who use such Helplines remain untreated and unhelped.

- What is the role of prevention and support for potential offenders? Organisations like ‘The Lucy Faithfull Foundation’ in the United Kingdom provide access to appropriate sources of support and intervention to help potential offenders manage their thoughts and behaviour and to assist them to live responsible, fulfilling and law-abiding lives.

Prevention needs to have a greater public policy and political commitment than it has enjoyed up to yet.  The evidence base for treating sexual offending as a Public Health issue is now robust enough to support such a change. The issue is, who is brave enough to lead such a change in attitudes, responses and policy?

- Is the legal framework adequate and up-to-date to deal with the various aspects of the problem, for instance in relation to young people (children) sending or possessing sexually explicit images of themselves or their partner?

I am not sure about our legal framework in this respect.  We are generally punitive and exclusionary, especially about sexual offenders including young ones.  Sentencing, despite some claims to the contrary, is not particularly well designed to rehabilitate, change or treat people.  Part of the issue here is about what we think is a crime, acceptable or unacceptable behaviour, and acceptable or unacceptable risks and harms.  This differs across society and is changeable over time: ‘sexting’ is a good example of this.


I am wondering whether the estimates of the number of websites with child sexual abuse content (31,266) provided here include websites located not on the traditional internet but rather on the Darknet? If not, the problem is much bigger than these estimates indicate. With the move to the Darknet, the monitoring, prevention and control strategies that will be needed have to change. New technology = new criminal opportunity.


On the Dark Web, contributors may be interested in a one-day event entitled: "Glimpse into the Dark Web: Understanding Online predators, grooming, sexualisation and radicalisation” information at: https://ineqe.com/glimpse/ with key speakers including Jamie Bartlett author of "Dark Net".


Thank-you IWF for introducing this important topic. As with other types of sexual abuse, online or internet offenders as a term describes a host of behaviours - the abuse, filming, posting, dissemination, access, downloading, sharing etc. I suspect that there are different, albeit overlapping motivations and risks associated with these constellations of behaviour, which as has been pointed out, requires quite a diverse response. This is not just about treating the individual who has accessed and viewed the images, or even prevention of future contact offenders (although as DMU points out this is certainly the risk we see and thus can intervene upon). But as UoML pointed out finding and removing images, and obviously preventing/discouraging access in the first place are means of disrupting the market for such images (this has its own challenges).

In Canada, Michael Seto has developed the CPORT - a risk tool for individuals accessing child sexual abuse images. Although not used clinically yet, it shows great promise as an actuarial instrument for distinguishing risks among this group of offenders. Including items about the offenders, the nature of the images, etc.

I'm wondering if we are also considering individuals who use the internet for sexual interaction with young people - such as chats or message boards. Some are certainly exploring means of accessing victims for contact offences, but there appears to be another group who seek out and use exchanges for stimulation, with no apparent interest in further contact. However, we don't know much about these individuals, as unlike images on a hard drive, message posting etc. are usually not accessed or included as part of investigations.

McM 2

Generally, I think we can all agree that the advent of the internet was a monumental leap forward in our social evolution. The world wide web has contributed to greater understanding of so many things, and has definitely made the world a lot smaller. However, in our own inimitable fashion, Homo Sapiens can't seem to do anything amazing without finding a way to muck it up.

When I was Kurt Freund's assistant all those many, many years ago (context is important here), I was tasked with finding realistic stimuli for use in phallometric testing. This resulted in me making "friends" (so to speak) with members of Joint Forces Project "P" - a collaboration of the RCMP, Ontario Provincial Police, and Metro Toronto Police. They were the clearinghouse for all of the "pornographic" materials seized by law enforcement. We were most interested in the child abuse images and torture materials - for use in sexual psychodiagnostics. However, most things were in Super 8, Polaroid, or poorly transferred VHS. We came away with some usable stuff but, generally, most materials were of such poor quality that they were unusable - at least in a scientific/clinical milieu.

Scoot ahead some 25 years and pretty near anyone with an internet connection (even on a watch!!) can find and download high quality child abuse images and all sorts of other nefarious and deviant materials, in a heartbeat. I'm pretty sure this is not what Al Gore envisaged when he "created" the internet (insert smiley face here). There is so much pain and suffering on the web that it screams out for better controls and monitoring - but, at what costs?

The ability and freedom to post, peruse, and comment on virtually any topic with virtually any person anytime and anywhere in the world is a miracle of human ingenuity.

I'm very, very interested to see where this discussion goes, as we debate such fundamental human rights as freedom and safety in the vast and often controversial sphere of the world wide web. How do we achieve both interests? Can we achieve both interests?

Thanks for your challenging questions, IWF. Let's see what we do with this...


Based on previous posts, here are some broad statements regarding risk to reoffend of users of child sexual abuse material (I am happy to share the referenced studies):

1.       Users of child sexual abuse material (CSEM) are a low risk for future contact sexual offences (and especially with only CSEM offences; see Babchishin, Hanson, & Vanzuylen, 2015: Seto, Hanson & Babchishin, 2011).

2.       Despite low base rate of contact sexual reoffending (about 0.2% for CSEM only offenders [that is no contact sex offence, excludes those who committed a contact sex offence to create CSEM] and 6% for those with both CSEM and contact offences after 5 years; Graf & Dittmann, 2011), these offenders do differ on their risk to commit a contact sexual offence. Luckily, risk factors have been identified to rank order child pornography offenders in terms of risk to reoffend (Seto, 2013; Seto & Eke, 2015). (Mixed offenders could be up to half of convicted users of CSEM; Seto et al., 2011).

3.       The major risk factors for any sexual recidivism among CSEM users (these studies included both CSEM only offenders and mixed offenders with both CSEM and contact offences) are related to general criminality, such as young age, prior violent offenses, juvenile record, and prior criminal history (Seto & Eke, 2015). The content of the child pornography collection also provides information about the likelihood of recidivism; specifically, the ratio of boy to girl content (Eke & Seto, 2012: Seto & Eke, 2015). Opportunity also matters. Motivated offenders with access to children are most likely to sexually assault children whereas pedophilic offenders who frequently use computers are most likely to view child pornography (Babchishin et al., 2015).

4.       Static risk tools have been modified (e.g., the Risk Matrix-2000), but do not do a good job at rank orders CSEM only offenders (available studies suggest they do well with the mixed offenders, those with both CSEM and contact sex offences).

5.       Dynamic risk tools (e.g., STABLE-2000) seem to do a much better job for CSEM-only offenders. For mixed offenders, risk tools do well for rank ordering offenders. Indeed, risk tools do not need to be modified for mixed offenders with an official history of a contact sexual offence (e.g., Static-99R)

6.       The CPORT was found to predict any sexual recidivism, contact sexual recidivism, and any recidivism after the 5-year follow-up. Unfortunately, although it predicted for the subgroup of child pornography-only offenders, it did a poorer job with mixed offenders (AUC of .69 vs. AUC of .80). Also, the CPORT did not predict recidivism among child pornography offenders who had no other criminal history (i.e., first-time offenders). Currently, the CPORT seems to be a promising tool, but requires cross-validation. The authors have drafted coding rules for interested clinicians.

7.       For CSEM only offenders, dynamic risk tools may provide more promising results (given it is not dependent on criminal history variables for which many CSEM users may be missing). We need more research on that front.

So who would be at risk for cross-over offences (i.e., going from a CSEM to a mixed offender)?  The literature suggests that those that pose a higher risk for cross over offences would be expected to have high levels of pedophilia, high levels of antisociality, have access to children, and have few psychological barriers to committing a sexual offence. Conversely, CSEM offenders would be expected to be low risk for contact sex offences if they score low on measures of general antisociality, have limited access to children, and have psychological barriers to committing contact sexual offences. As a default, CSEM-only offenders should be assumed to be very low risk to reoffend with a contact sexual offence. Proven risk factors (e.g., greater antisocial traits, greater criminal history, younger age) would be needed to justify a higher risk for a child pornography-only offender. This subsample of CSEM users could be targeted for alternative responses (IWF and posts).


Just some supplementary comments on UoO's response to IWF's questions:

1) The requirement for prioritization of online offenders for intervention is not ideal, but in the current climate, with many thousands of offenders in online circulation and readily identifiable to law enforcement, it is absolutely necessary. UoO has already rehearsed the requirement for prioritization at the level of law enforcement.  To my mind, effective prioritization of those more likely to contact offend / with recidivistic potential at this juncture is critical, as it is largely here that important, first-time decisions are made to select offenders for criminal justice intervention and latter inclusion within secondary prevention systems. This factor, coupled with limited resources, a paucity of alternative interventions, etc. makes effective investigative targeting of such cases all the more imperative. 

I think it's essential to prioritize higher-risk offenders and to make other informed policy decisions, e.g., prioritizing production cases (which necessarily involve exploitation/abuse of a child) and active distribution (traders) over passive distribution (e.g., making files available on peer-to-peer networks) and receipt/possession alone.

Absolutely. Re the identification of production cases (and the role of technology in supporting same), I would just like to mention a software I was involved in developing - the iCOP toolkit - developed with European Commission funding support. Unlike other technologies which rely on databases of known CSE images/video to identify new cases of CSEM offending, this technology allows for the identification of new CSE material that has not previously been seized or hashed using solutions like Photo DNA. This ability allows law enforcement to identify new images/video not yet known to them, thereby supporting the identification of production cases. You can read more about it here: http://ieeexplore.ieee.org/stamp/stamp.jsp?arnumber=6957295

I absolutely agree with the point re the prioritization of active vs. passive traders of CSEM. UoO's point speaks to a more fundamental need to better understand the mediating influence of technology on online offending behavior and to accommodate this mediating influence in risk profiling or assessment with online offender populations. At the level of investigating law enforcement, there is a definite requirement for more discriminating risk profiling (not "risk assessment" :) ) systems that allow for the identification of high risk cases in online investigations. A critical feature of any risk profiling in online investigation is the identification of volitional behaviors, where the behavior you see online reflects the actual motivation or intent of the user. This is not easy with online behaviours in CSEM users. Very often the technology used by the offender to enable their trading or other offending interferes with the behavioral presentation and can undermine the quality of the assumptions investigators (and others) can make about the motivation/intent of the offender. For example, users of Peer to peer systems may download many CSEM files from Peer to peer, but this 'behavior' tends to be affected by the design of P2P systems, which require users to download files based on a reading of their filename before they can be viewed. Therefore users may download many files of different sexual genres in order to locate material that appeals to their actual sexual interest, with the effect that their sexual interest is only in a subset of what they actually download. This affects the quality of the assumptions we can make about an individual offender's sexual deviance / risk based on the number or type of downloaded files we find in their possession. We urgently need more empirical work to understand the mediating influence of technology on online behavior, to help us to differentiate online behaviors that can be reliably used for profiling and risk appraisal, and thus build better case prioritization systems for investigative use.

Regarding other risk profiling/case prioritization tools for use in law enforcement investigations: Under the iCOP project we have developed a risk profiling methodology for use by investigators that profiles sexual deviance (paraphilia) in users of Peer to peer systems, as a basis for case prioritization decisions. Here, CSEM offenders on Peer to peer systems are prioritized to the online investigator based on the identification of a problematic profile of sexual deviance (paraphilic interest) in their peer to peer searches. The search behaviors of P2P users (i.e. user searches using CSEM related keywords) proved to be a useful behavioral indicator for risk profiling purposes, as it is here that the user directly betrays their sexual interests and motivations. Once validated, this system will be embedded within Peer to peer monitoring technologies used by investigating law enforcement as a case prioritization support.

3) Another Thorn project uses decoy files placed on known sites/networks for distributing child exploitation content. Instead of seeing the image, the user instead gets a warning reminding them that the behavior is illegal, that they could be tracked, and pointing them to resources from Stop It Now! to get help.

UoO and others, are you aware of any work to evaluate this, or similar initiatives?  I'm not making any assumptions around this Thorn-supported intervention (Thorn is an exemplary NGO, as are its projects) but there are other online initiatives like this (e.g. police/search engine use of splash pages or pop up warnings as deterrents/interventions) which may be helping, but we have insufficient empirical understanding of their efficacy vis a vis prevention or desistance. If you are aware of anything, it would be really interesting to learn more about it. I guess the point here is that we need better empirical evaluation of the efficacy of these online interventions for CSEM offence prevention/desistence.


Following up on the last comment by UoO, I have suggested in my work that policies and practices will need to evolve to respond to three overlapping groups of offenders: contact sex offenders against children, the group we know the most about; CSEM offenders, some of whom have committed contact offenses and/or online grooming offenses; and online grooming offenders, again some of whom may have committed contact or CSEM offenses.

There are differences in pedophilia, antisociality, opportunity (access to children) and other psychological factors. For example, few online grooming offenders are likely to be pedophiles given they are mostly targeting young adolescents.

Responding to the original questions from IWF:

(1) How do we deal with the volume of people accessing online child sexual abuse material? Can we (should we?) prioritise high risk offenders within the wider group of offenders? Are existing threat assessments adequate? Is there enough information/research regarding the potential transition from online to offline abuse?

To me, the sheer numbers of individuals involved makes it clear that a purely criminal justice response (arrest & prosecution) is not enough on its own. As I heard a high-ranking prosecutor say in the US (a nation that is not shy of a heavy hand in criminal justice): "We can't arrest our way out of this problem."

I think it's essential to prioritize higher-risk offenders and to make other informed policy decisions, e.g., prioritizing production cases (which necessarily involve exploitation/abuse of a child) and active distribution (traders) over passive distribution (e.g., making files available on peer-to-peer networks) and receipt/possession alone.

Following up on a comment by DMU, I would emphasize that the KIRAT is *not* a risk assessment tool in the sense that I understand it, despite it being part of the name. The KIRAT identifies factors that distinguish online-only offenders from dual offenders who have committed both online and contact offenses *at point of police contact*. That is useful in prioritizing cases that involve current or past contact offending, but it is not the same as determining who is at greater risk of contact offending in the future (though it is true that dual offenders are higher risk than online-only offenders). As we reported in our 2011 meta-analysis, the sexual contact recidivism rate of online offenders is lower than for contact offenders, but of course some do go on to sexually reoffend, either with another online offense or a contact offense.

To assess risk of contact offending in the future, we need a prospective risk assessment tool. There is some preliminary evidence for the use of a modified Risk Matrix 2000, a standard risk assessment tool in the UK (but not Canada or the United States) and for the Child Pornography Offender Risk Tool (validated in Canada only so far).

 (2) What should the response be towards the offenders? Is it possible to arrest all offenders or should alternative initiatives be explored? And if so, what could these initiatives look like?

The criminal justice system does not have the capacity to investigate, much less arrest and prosecute and sentence, all online offenders. To me, this is a strong argument for investing more in prevention and for efforts to reduce demand. There is a strong analogy, to me, between the current "war on online offending" and the failed "war on drugs" in terms of prioritizing production/distribution over possession and in terms of recognizing that treatment and criminal justice alternatives were needed for users.

(3) What is the role of prevention and support for potential offenders? Organisations like ‘The Lucy Faithfull Foundation’ in the United Kingdom provide access to appropriate sources of support and intervention to help potential offenders manage their thoughts and behaviour and to assist them to live responsible, fulfilling and law-abiding lives.

I think the prevention efforts of organizations such as Stop It Now!, Dunkelfeld, the Lucy Faithfull (e.g., it's revamped croga.org website) are important avenues to pursue.

I also believe that, just as technology has facilitated the access and exchange of child exploitation materials (child pornography in legal terms), technology can also play a role in intervention. I am the pro bono scientific advisor to a non-profit organization based in the US, the Thorn Foundation (previously DNA Foundation) that was founded by actors Demi Moore and Ashton Kutcher to pursue technological responses to the problem of online sexual exploitation.

The Thorn Foundation has been involved in efforts such as the adoption of photoDNA to automatically check uploaded images to known child exploitation images from a database maintained by the National Centre for Missing and Exploited Children. This technology was developed by Microsoft in collaboration with Boston College and is now used by major platforms such as Facebook. Uploading a known child exploitation results in suspension of an account and notification of police. Wide-spread adoption of this technology -- and new technology that is being worked on to do the same for video -- could have a dramatic effect on availability of content.

Another Thorn project uses decoy files placed on known sites/networks for distributing child exploitation content. Instead of seeing the image, the user instead gets a warning reminding them that the behavior is illegal, that they could be tracked, and pointing them to resources from Stop It Now! to get help.


I agree that the scale and nature of the sharing as well as downloading of Child Sexual Abuse imagery has up-scaled immensely since the advent of the internet, the real question [for me], building on IWFs and other comments here, is how we respond to it? I think that looking at downloading Child Sexual Abuse imagery as a potential warning sign and opportunity for an early intervention [secondary prevention, if you will], but the question is how we facilitate that - through a health response [programmes, doctors referrals, community nursing, online treatment/management programmes], a criminal justice response or a combination of both? Who will reach out and what will that look like? I think that there is a real opportunity to do something proactive here.

In respect to the 'dark web', I think that this brings us back to the roles and responsibilities of governments as well as internet service providers. Given the international complexity of running and managing the internet [epically in terms of where material is hosted and routed through] we have to get better at multi-national/multi-corporation working and agreements. But who should set up and intervene, the UN, multi-national corporations. This is such a challenging issue, I think that it has more to do with international relations and politics than child protection [as harsh as that sounds].

My question is - how do we reach out and what is the best way to do that?


I am of the view that technology has helped create this problem of online sexual exploitation and abuse, and technology is needed to respond effectively to this problem. I don't think technology alone will do it, and have argued for a multi-prong approach that includes technology, prevention efforts through the health and social service systems, as well as criminal justice. But the challenges for health, social and criminal justice is scale. Even with massive increases in funding -- unlikely but one can always hope -- online services, community programs etc. can only reach a small fraction of perpetrators, and even then they will mostly reach those who are already motivated to seek help. There are many others who won't be reached, who aren't motivated to seek help or are interested but afraid to identify themselves in even the slightest way.

Part of the technology response involves the complexities of engaging in international law, national policies, and the big corporations. I don't think we have to re-invent the wheel here as the international communities have managed to get their act together -- to some extent -- in responding to previous global issues such as climate change.

Here's an example of what I mean about technology and its potential for scale and reach:

photoDNA is a software program created by coders at Boston College with the support of Microsoft. It can automatically compare any image, without human involvement, against the database of known child exploitation images kept by the National Centre for Missing and Exploited Children in the US. The software goes beyond simple hash matching because it can detect matches even if users change file properties such as size, filename, cropping etc.

This software is freely available and has been adopted by Facebook and other major companies. This means anyone trying to upload known child exploitation images to Facebook -- one sneaky way of sharing large batches of images -- will have that account suspended and the pertinent information sent to NCMEC.

Imagine the impact on access and distribution of child exploitation images if every photo-sharing site -- or even better, all major ISPs -- implemented photoDNA. It won't catch the uploading and access of previously unknown images, but analyses of collections seized from perpetrators tells us that the majority of content is already known to NCMEC.

photoDNA could be promoted as "corporate responsibility" across photo-sharing sites and ISPs; it could be expanded by linking (overlapping) databases of known images kept by the RCMP in Canada, CEOP in the UK, Interpol etc. I know that Microsoft is working on a videoDNA (much harder tech problem). 

I think of photoDNA and similar software as the equivalent of situational crime prevention. Irrespective of perpetrator motivation, online or offline prevention and intervention resources, policing etc, wide implementation of this software could have a really big impact.

McM 3

Thank you very much for giving the opportunity to participate to this discussion.

Regarding the aspects of policies and practices mentioned by UoO, I find that the law does not always fit into the clinical finding. As an exemple, in the French law, if someone is charged for possessing child pornography but can prove that the teenager on the image is in fact an adult dressed as a minor, the individual will not be convicted (of course this is certainly difficult to prove anything because of the fact that the images travel around the world).

McM 1

I agree with UWE that because of the nature of the internet a multi-national approach is needed. Where the abuse occurs, where it is uploaded, and where it is downloaded could easily cross multiple jurisdictions. And unless there is a coordinated investigation with appropriate sharing and access we will always see fragmented management.

My other thought is that more needs to be done at the uploading stage. There has been lots of attention to how to assess, treat and manage those who possess the images, I think because that is often who is easiest to detect and prosecute. However, the source abuse and uploading continues with impunity. Governments put lots of resources into developing technology for monitoring other types of criminal activity or threats to state. Perhaps this could be adapted for file sharing sites etc.

There has been discussion on the ATSA listserve about peer to peer (youth) sharing of sexual images of themselves and how this differs or ought to differ from child sexual abuse images. Just as the response for peer sexual behaviour does not come from the courts, but from parents or other responsible adults, should CSAI not be distinguished the same way?


This is already really interesting, so thanks for including me.

This is a monumental problem for all countries and I'm afraid I don't have any solutions. A police officer told me a few months ago that new technology in the UK had enabled the police to track over 25,000 potential internet offenders in the UK. The police are currently doing nothing about it as they simply don't have the resources - this applies to police, court time and prison spaces. We clearly therefore need to prioritise who we concentrate our scarce resources on. I agree with others that we should focus on those that make and then distribute but as we all know this is not easy, especially when we are talking about the WORLD wide web.

I think we should differentiate between imagery of child abuse and young people sexting. While there are problems with the latter I certainly don't class these people as sex offenders and would not in any way want them labelled as such. For this we need to educate our children about the risks and dangers of this type of activity.


I think this point is worth emphasizing, this is not a problem that police, community programs etc can fully address because of scale. I recognize that criminal justice, health and social responses are also needed, and ideally coordinated, but I think the scale of the problem -- a result of online technologies -- requires a similarly scalable tech response.


I have read the posts with great interest, and followed the debate that is developing. UoO and UoO have responded in detail regarding the current (empirical/theoretical) knowledge concerning CSEM users, especially concerning their risk levels; in response to this, I like to share some of the (controversial?) questions that are currently on the research agenda.

1. Through working with the UK police and community treatment centres, such as the Lucy Faithfull Foundation, I can see the systemic consequences following from the large numbers of online offenders; especially with regards to policing, this requires an attention and resource demand that is very difficult to meet, and my feeling is that there is a (a slow but steady) move towards early interventions, maybe one day similar to the model proposed by the Prevention Project Dunkelfeld in Germany who offer confidential advice to individuals with a self-identified interest in children. We had a debate at the NOTA conference two years ago if such a model could/would work in the UK - and a very strong discussion emerged. It would be great to know if other countries have similar/ different prevention models and how others in this network consider their value and societal impact.

2. To qUoOe the previous post: "Users of child sexual abuse material (CSEM) are a low risk for future contact sexual offences (and especially with only CSEM offences; see Babchishin, Hanson, & Vanzuylen, 2015: Seto, Hanson & Babchishin, 2011)" - could this be understood as an argument to introduce a less resource-demanding penalty (such as fines) rather than treatment for CSEM users, and to focus these resources on reoffenders? The psychological literature has been challenged as to "neglecting the victim perspective" and I wonder how my colleagues combine these positions?

3. How should the use of CSEM be considered in child custody evaluations? There is a debate in the psychological literature if and how the use of CSEM is linked to a sexual interest in children, and empirically, there appears to be a low risk of contact sex offending. On the other hand, there are questions about the well-being of a child under supervision of a parent who has engaged with CSEM, and potential risk escalation linked to its use. Has anyone got experience how a judge/ psychological assessor currently evaluates the use of CSEM in legal decision-making concerning unsupervised access to children?



#1. I know there is interest in the Dunkelfeld approach, for example my hospital's sexual behaviour clinic is introducing and evaluating a prevention program for self-identified pedophiles and hebephiles. I'm aware of interest in other jurisdictions as well but I don't know how far they have come.

#2. There is always a tension between risk to reoffend and other purposes of criminal justice responses. From a purely risk and resource perspective, imprisoning CSEM possession offenders is not efficient. At the same time, victims and the public are calling for bigger penalties and a "stronger" response. One potential road forward is a restorative justice approach.

#3. I don't know how much use of CSEM factors into custody and access evaluations and decisions. My hunch is that it would have a big impact, even though the empirical evidence is that CSEM users pose a low risk of *detected* contact sexual offending. In the 2011 meta-analysis we published in SAJRT, we found a big gap between criminal record (1 in 8) and record plus self-report (1 in 2) so it is probable that a sizable proportion of parents who have used CSEM have in fact committed undetected contact offenses.


Thank you for all contributions to this interesting debate. I would like to take this opportunity to address some of the issues raised by participants from an ‘operational’ point of view.

-          Dark Web: The number of webpages mentioned in the opening statement (31,226) relates to the actual number of webpages assessed by the IWF in 2014 as depicting child sexual abuse. This is not an estimate of the total number of webpages that exist depicting this content but merely what the IWF encountered in one year of operations. The numbers for 2015 will be released soon and will show a significant increase. The IWF encountered 51 instances of links to CSAM on hidden services in 2014. Most often the content was still hosted on the open web and all instances were reported to law enforcement. So the IWF hasn’t seen a huge increase in the number of reports related to the dark web – but of course we only have data on what we encounter. The International Association of Hotlines (INHOPE) is due to release their annual report soon and this should provide us with a more global overview of the amount of CSAM encountered by hotlines.

-          International response to combating online CSAM: Currently, the international cooperation for hotlines to remove content at source is as follows. If a hotline encounters content which is traced to another country (for the UK, this is the case in 99.7% of the CSAM encountered by the IWF) it will pass on this information to the hotline in the ‘hosting country’ using the INHOPE reporting mechanism. The hotline in the ‘hosting country’ will then process the report using its own procedures and will liaise with national/local police forces. If there is no INHOPE hotline in the ‘hosting country’ reports are passed on to local law enforcement. Besides removal at source, the IWF also provides services to the online industry (URL List, Hash List, Keyword List,…) which companies can use in the various markets they operate in.

-          Self-generated content/sexting: The IWF increasingly sees sexually explicit content which could be assessed as ‘self-generated’, using cameras on mobile phones or webcams. For older teens, this would typically be content that is sent to boy or girlfriend. For younger children, this is more often related to some form of coercion by people on the other end of the connection. Unfortunately, this material ends up on the open web and in most cases (88% of the content encountered in an IWF study of 2012 – accessible on iwf.org.uk) it is taken from the original upload location and brought together on websites which have ‘collections’ of these images.

McM 3

I see the rational for UoL's second point, regarding less resources-demanding intervention. It occurred to me when I assessed some of these offenders to feel that a formal psychotherapeutic follow-up may not have much impact. However, I also had the impression that if we do not have enough resources to put on this population, we may miss the child pornography consumers who are also a hand-on offenders but unknown to the police at the time of the assessment/treatment.

There are some hands-on sexual offenders who stop assaulting children when they access child pornography online. I was wondering if this means that this population is really clinically different from the one who keeps offending. Do we know which psychological factors may influence the change in the behaviour (insight, impulsivity) and should we have specific therapeutic approach for them...

Does anyone have some thoughts or data about it ?

McM 2

So, if we follow from comments made by the lot of us...

I don't pretend to have any real sense of the magnitude of the issue. Like most, I know that it's there and that there are increasing numbers of offenders coming into the system. As a Static-99R trainer, the most common question I get in the US is, "What can we use on our CSEM offenders, if we can't use the Static?"

UoO's qUoOe from the police officer is telling: There are likely too many online offenders to "arrest our way out of this". It would also seem that to principally target the downloaders (or users) would be akin to picking the low hanging fruit. In my view, the producers and distributors (over and above simple sharing by users to obtain more materials - like webmasters) of the materials should be our focus. If those people who are strictly downloaders are at relatively low risk, perhaps, some different way of responding to their behavior is required. At present, many downloaders - when caught - end up with (sometimes) inordinately high sentences in comparison to the potential risk they pose.

Having methods and tools that would help us to triage the users most likely to be a continued problem would be welcome, and it sounds like there are efforts to do something of this sort. But, again, is that where our strongest systemic efforts should be directed? Our response could be something along the lines of what some jurisdictions have done with drug courts. Recognizing that the users have a "clinical" issue - over and above having broken the law - we could offer psychological assistance, while directing our major criminal justice efforts to holding accountable the producers and distributors.


I think the analogy to the so-called war on drugs is apt, I've made the same point in presentations.

The "war on online sexual exploitation and abuse" is not going to succeed if criminal justice and other sources are focused on arresting end users (accessers & possessors) rather than providing prevention and early intervention resources and focusing police investigators on the traders and producers. And, like the war on drugs, we won't make gains without addressing demand as much, if not more, than supply.


This has been a very interesting discussion. I think most of us agree that CSEM is a public health problem that requires more than just a criminal justice response. There seems to be some agreement that we should prioritize ‘higher risk ‘ CSEM offenders for police/criminal justice interventions, but that lower risk offenders could be transitioned to non-criminal interventions, such as treatment and restorative justice courts.

To do this requires a method of case prioritization that is empirically justified. This is where researchers can help: identifying factors amongst CSEM that are related to contact sex offending, and then creating and validating a prioritization instrument. The work by Michael Seto and Angela Eke are helpful in regards to factors associated with reoffending with contact sexual offence. There is also the KIRAT that can be used to identify CSEM users that are possible mixed offenders (also having a contact sexual offence).  With case prioritization, very low risk offenders could be targeted for an alternative intervention. Treatment, however, is difficult because of mandatory reporting laws (some states have/are passing legislations that would make CSEM use meet mandatory reporting requirements). That said, there are also online treatment programs that could be helpful (e.g. Stop it now, http://www.stopitnow.org ) until Dunkelfeld type programs are available.

A related point: mandatory minimums are lengthy and are typically not proportional to the risk posed by CSEM offenders. This means we are using limited resources to incarcerate a group of offenders that are mostly at a very low risk to reoffending, rather than focusing on incarcerating individuals who pose a risk to the community. Public safety is not the sole purpose of criminal justice responses (e.g., punishment), and so a restorative justice approach could be one-way forward as UoO had suggested. Indeed, similar to drug-treatment courts and domestic violence courts, perhaps we will see the use of CSEM courts in the future (e.g., McM 2 point). This would require courts and politicians to see the benefits of an alternative approach for some CSEM users (knowledge translation efforts would be needed).

Lastly, there was also mention of primary prevention programs (e.g., UoL). For example, targeting youth to ensure they understand healthy sexual boundaries. Creating online or in person help for individuals who are concerned with their sexual interests. Increasing public support for prevention is important, and necessitates that the public understand that child sexual abuse is indeed preventable. Public and politicians may then be more apt to funds such prevention programs (e.g., work by Elizabeth Letourneau).


I totally agree with your last point McM 2 in the sense that perhaps we should see the downloaders as having a clinical issue. As we know viewing all forms of child sexual abuse imagery are illegal including those which don't depict real children. This means that for those who have a sexual preference for children there is no legal outlet for them. I'm not necessarily advocating for decriminalisation of this activity but I do wonder whether we need to treat the need rather than punish the activity. This would obviously be dependent on the type and scale of the pictures that were involved.

There is an interesting article posted by BBC News today which has the latest IWF figures:http://www.bbc.co.uk/news/technology-36091424


This sounds like an interesting proposal - and one that agrees with the current philosophy that is becoming more widespread, e.g., by Dunkelfeld or the approach UoO describes in his clinic.

However, looking at the initial desistance rates reported by Dunkelfeld (http://www.ncbi.nlm.nih.gov/pubmed/25471337), their confidential intervention didn't have the effect that was hoped for. One of the CSEM users we interviewed stated that the day of their arrest was both "the worst and the best day" of their life, as for the first time their online behaviour was "out in the open", and they were "forced" to deal with their offending. Of course, differently to the German approach, this was - at the time- also the first time for them to be offered access to support avenues. Many of our interviewees described a cognitive detachment from their online behaviour as "not being real" (Lucy Faithfull uses the term "bubble"); which was also identified when we investigated the implicit theories of CSEM users (http://eprints.lincoln.ac.uk/19676/). So I wonder if the low reconviction rate linked to CSEM users could also be linked to the experience of "making it real" through having to deal with offender management services - and that this could be potentially missed in the confidential interventions currently provided.

An excellent approach would be one that allows us to deal with the large number of offenders, initially based on their self-management and providing support avenues pre-arrest (and thus, potentially without too much interruption of their lives to maintain social integration and support), however at the same time focusing on the "realness" of online behaviour and the consequences linked to this type of offending.


Although I am not naïve I was surprised and shocked by the recent IWF figures, they really reinforce in me that we are not truly aware of the extent and nature of CSEM online; I am beginning to wonder how, if ever, we will be?

The scale, type and nature of the material does raise the legitimate question of how we respond to online CSEM, grooming and online exploitation; especially given that a sole criminal justice approach does not seem to be working. I too agree that we should consider using low level users of CSEM as an access point to treatment and prevention, rather than sticking to a complete criminal justice approach. Identifying these low level CSEM users and redirecting them to treatment and support is a worthwhile secondary prevention tool (I recognize that viewing does not always lead to doing). You can imagine a scenario where they get contacted by the police and a health professional to discuss their online activity and how to respond to it. In addition, how does this translate into youth with a sexual interest in other youth [or younger], peers looking for [what in their instance] age appropriate material and individuals with Learning Difficulties...

I think the question might be what is the role for the individual, community and state [especially via schools, health care, etc] to foster a conversation about this and the ways to get a grip on the situation? Especially when [as said on Wednesday at the Leverhulme conference in Bristol] that people are criticizing the IWF for trying to censor and control the internet by tagging/removing issues.


UWE - I think the biggest learning from Dunkelfeld was that a public/open approach IS possible about sexual interest in children - their public advertisement campaign (https://www.kein-taeter-werden.de/) is an example of lowering the public threshold and making it possible to raise one's concerns, issues, calls for help (...) at an earlier stage.

UWE also raised the issue of earlier interventions in school - we need to raise the bar with our young people. Just to veer out a little: It is still possible in some schools for parents to opt out of sex education - but we need to prepare young people, both in terms of the sexual exploration the internet provides but also the sexual victimisation that can occur online. Just as the internet has now become a healthy component of the school curriculum (in terms of access to resources, to teach programming skills etc.), it also provides young people with a constant reinforcement schedule that they need to self-manage; this is a huge self-management task that children need to acquire, which places new developmental challenges. To me, the most interesting finding from http://www.mdx.ac.uk/__data/assets/pdf_file/0026/48545/Basicallypor... was the reluctance of schools to "talk about it".


UoL & UWE - just a brief observation regarding young people and the importance of education as a preventative mechanism. One of the core issues that is becoming apparent to those working in schools with children on these matters (certainly in the UK and Ireland) is that the boundary between sex/relationship education and school-based 'online safety' interventions is becoming increasingly difficult to draw. One view is that sex and relationship education curricula need to begin to comprehensively address issues around technological expressions of sexual behavior so that young people can better understand and manage their own sexualised use of technology. For example, this might involve increasing young peoples' awareness of potential issues around problematic pornography consumption, or the implications of posting self-generated sexual materials in publicly accessible spaces (e.g. image harvesting and reposting on 'para sites' as described in the 2012 IWF report), the legal/social consequences of sharing sexual images of other young people online, associated issues around sexual consent, etc. Interesting UK perspective from Prof. Andy Phippen of Plymouth University here: http://blog.politics.ox.ac.uk/digital-rights-and-pornography-a-chil... ;


We would like to thank all participants for a very detailed and rich discussion.

As the internet evolves, the problem evolves and so do the possible solutions to the problem. It was very interesting to hear the various approaches, often originating from different contexts.

Even an attempt to summarise the debate would do injustice to the full extent and detail of the discussion.  As such, I will only mention a couple of general points.

Over the past couple of weeks, we discussed the nature of the problem, distinguishing between various types of offending (producing, distributing and viewing CSAM), the technological tools available to combat the availability of CSAM and the responses in terms of prevention, treatment and criminal justice.

One of the reoccurring difficulties for dealing with online CSAM and CSE touched upon in this debate was ‘scale’.

The discussion focussed on the possibility to distinguish between various risk levels and which approach should be taken for these different categories – as well as who could be involved in addressing the problem (i.e. responsibilities of the individual, state and society).

The debate was also rich in references and links to further information. I hope these references and the content of the debate itself was of use to all participants.

From our perspective, the debate provided an extremely interesting addition to the work we do in our hotline and we look forward to continue following the various initiatives in this field. 

Again thank you for your contributions and special thanks to the organisers for facilitating this discussion.



Previous Article
Next Article
ATSA | 4900 SW Griffith Drive, Suite 274, Beaverton, OR 97005 | Contact Us

To ensure delivery, please add 'sarah@atsa.com' and 'sarah@atsa.mmsend.com' to your email address book. If you are still having problems receiving our emails, see our whitelisting page for more details.

Click here to unsubscribe
Members Menu ATSA Journal ATSA Conference Update Your Information Home