Using Forensics to Tackle Child Exploitation

By Sarah MooreReviewed by Emily Henderson, B.Sc.

The invention of the internet is hailed as one of the greatest technological advancements in history; however, it is not without its dark side. The internet has allowed us to be connected, to communicate and share with others around the world, and to do so with anonymity, if so desired.

The internet, unfortunately, has been used as a tool to facilitate sexual abuse offenses against children, and it is believed that the internet has increased the manner and speed at which indecent images of children are produced and exchanged.

Image Credit: HTWE/Shutterstock.com

In general, sexual abuse offenses against children are viewed as the most heinous crimes. This type of offense is escalating, partly due to the ease of access to images and sharing that the internet allows. In 2019, the UK saw a tenfold increase compared with the previous 3.5 years of arrests and the number of children safeguarded or protected in connection with sexual abuse offenses against children. The National Center for Missing and Exploited Children in the US received 16.9 million reports on online material depicting child sexual exploitation.

There is an urgent need to protect children from exploitation. Police are under increasing pressure to identify the perpetrators of these crimes and to stop them from committing further offenses. Each time an image is made, a child has been exploited. And every time that image is shared and accessed, that child is exploited again.

Law enforcement agencies recognize the need to remove images of child sexual abuse from the internet. They must also catch the predators responsible for creating and sharing the images. Thankfully, digital forensics and AI are being leveraged to help.

Removing images of abuse from the internet with AI

In the UK, police use deep learning to recognize and categorize images of child abuse. Traditionally, numerous officers are required to view and manually categorize images, which is time-consuming and psychologically challenging. The new AI that has been developed can automatically recognize images of abuse and categorize them in terms of severity.

This technology is helping images to be removed from the internet. In 2018, 105,047 URLs hosting child abuse content were taken down. Since 1996, the UK-based charity the Internet Watch Foundation (IWF) has removed 477,595 pages containing images of abuse from the internet today. The rapid increase in the rate of removals we have seen in recent years can be attributed to better detection systems that recognize images of abuse more rapidly.

Image Credit: Andrey_Popov/Shutterstock.com

Catching predators

Not only is digital forensics helping to remove images of abuse from the internet, preventing them from being shared further, but it is also helping to catch the people responsible for creating these images. The Child Abuse Image Database (CAID) already contains millions of images. New images are compared against these to identify whether a child or geographic area where the image was taken has already been linked with an abusive crime.

In recent years, more digital forensics methods have emerged to identify perpetrators of abuse from images. Working with the UK Metropolitan Police in 2006, Sue Black, a professor of anatomy and forensic anthropology at the University of Dundee and director of its Centre for Anatomy & Human Identification, developed a unique way of using images to confirm an identity. Black was working on a case where a child had reported being molested by her father. When her mother didn’t believe her, she set up a webcam in her room. The webcam’s infrared mode caught the arm of the perpetrator. Black recognized that deoxygenated blood in veins showed up as black lines under infrared light, allowing the pattern of prominent veins to be used to identify the father.

Black has since used unique markings of the hands and forearms visible in abuse images to catch pedophiles. In 2009, Black spotted an unusually shaped lunula (found at the base of the nail) that was used to catch one of the ringleaders of Scotland’s biggest pedophile gang.

When combined with AI and deep learning, information about distinct hand and forearm markings have great potential to help identify and incarcerate perpetrators of child sexual abuse. This technology is still relatively new in terms of its application to fighting child exploitation, and it is possible that in the coming years, the technology will be developed further to enhance its capabilities in this field. Additionally, its development for use in tackling child exploitation is opening doors to leveraging digital forensics in other fields of criminal investigation. For example, it has been suggested that the AI trained to detect images of child sexual abuse could be trained to detect other crimes, such as violent crime. In this scenario, the technology could help speed up the analysis of CCTV captured at the time of a crime, which could lead to suspects being questioned earlier in the investigation.

Read Next: Digital Forensics in an Evolving World

Further Reading

• All Forensics Content
• Collecting Evidence at a Crime Scene
• Forensic Investigations of Natural Disasters
• Forensic Identification in Mass Disasters
• From Flower to Foul Play; An Insight into Forensic Botany