While there are many perspectives on the topic of crime, one opinion tends to be consistent around the world — the victimization of children is especially heinous. Thousands of trained professionals have dedicated their lives to combating crimes against children, but the task becomes more complex and challenging over time. Unfortunately, the same technologies that enable global trade and cooperation also enable criminals to have greater access to child victims than ever before in history. This problem has become widespread and serious enough to spawn its own category of investigation: Internet Crimes Against Children (ICAC).
According to Jason Webb, National Investigations Director for Voyager Labs and former criminal investigator with direct experience investigating ICAC cases, ICAC consists of a variety of related crimes, including (but not limited to) proliferation of child sexual abuse material (CSAM), solicitation of sexual pictures from a minor, manipulation of minors to extort money or sexual materials, and grooming of children for various types of sexual abuse. In the worst cases, the abusive activity moves from the internet to the real world, leading to sexual assault, abduction, human trafficking, and even murder. Fortunately, new AI technologies are being applied by investigators so that they can identify and stop ICAC. In most cases, this can be accomplished with information that’s readily available to the public…if the right technology is applied.
The Latest Digital Investigation Tools Are Turning the Tide
Many ICAC Task Forces are applying sophisticated technologies to assist online investigations in a variety of ways, using tools like “topic query searches” to find online activity related to crimes against children. For instance, these systems can find accounts trading CSAM, identify and locate the offenders using those accounts, identify potential victims, and compile crucial data to help law enforcement open a case against the criminals. This process is often remarkably fast. Webb indicated that the systems sometimes return usable results in seconds: “I was doing a demo and did a topic query and found an active page where they were trading CSAM pictures, and we were able to send the info to open a case.” Other investigators from Voyager Labs have reported the same pattern. The systems are so effective that they provide usable results, even when being used merely to demonstrate general applications.
A topic query lexicon is like a translation dictionary or code book. Experts who specialize in a certain type of criminal investigation (ICAC, human trafficking, gangs, narcotics, homicide) contribute their knowledge of criminal communications to build categories of current slang, terminology, and emoji codes so that investigation software can find hidden references to criminal activity online. As Webb points out, this allows investigators to contribute decades or centuries of combined law enforcement experience in a shared database, enabling various types of specialists to apply one another’s knowledge to their own research.
Investigative software can automatically search a variety of platforms throughout the worldwide web, deep web, and dark web, matching keywords from its lexicons to the communications it finds online. ICAC activities are often conducted on FB, Twitter, youtube, pinterest, Reddit, telegram, discord, and various platforms on the dark web, which these systems can analyze. Lexicons can also be put on monitors to alert investigators to matches, set to run daily, weekly, monthly, every few hours, or ongoing. Any time a match is found, it can email a clear, usable report to the designated recipient.
In many cases, all of this research can be accomplished solely through open-source intelligence (OSINT), which is available to the public with no warrant or special permissions. When sufficient evidence is gained to demonstrate that criminal activity is taking place, investigators can get social media search warrants to look more closely at the social media accounts where the criminal activity is taking place. The systems that identify these activities can be programmed to filter by times, region, networks, locations, dates, type of links, and activities.
One of the most useful aspects of ICAC investigation technology is network analysis. For instance, if there are pages where CSAM has been traded, AI can see who is interacting with those pages to do network analysis, identifying if multiple offenders are involved in ICAC rings. It is common for gangs to run ICAC activities and human trafficking, so these investigations have the potential to uncover criminal enterprises of complex organizations that work in a range of illegal activities. Once a criminal ring is discovered, investigators often follow the evidence to expose not only ICAC but human and drug trafficking, weapons trafficking, extortion, financial fraud, and violent crimes.
Some of these systems can get search returns that narrow down the activity geographically, clarifying the physical location of the criminals and the jurisdictions of related law enforcement entities. Webb explains, “If you are working and you get returns and filter by location, you can filter down to a specific area. You can take the info and contact the area law enforcement office for them to work the case.” This type of massive information-sharing and collaboration may be one of the biggest factors in reducing Internet Crimes Against Children because it empowers agencies that have this new technology to share their research with smaller departments that may not have the same resources.
In addition to identifying criminals, network analysis processes can also indicate potential victims, alerting investigators to children who have been targeted by criminals. Once law enforcement and families are alerted to the fact that criminals have attempted to contact a child, immediate steps can be taken to remove the child from potential harm. This, of course, is one of the best scenarios. While ICAC may never be completely eradicated, AI investigation software may be the key to dramatically reducing its devastating effects.
