These illicit activities form a massive global economy sector. While it remains hard to assess their full scope, transnational networked criminal activities are widely estimated to account for up to 5% of the world’s $80 trillion economy. These estimates put this illicit economy on par with, or even ahead of, the economies of most wealthy state members of the G7 group, lagging behind only the economies of the U.S., China, and Japan. Driven by such profit potential, these illicit activities are evolving: human trafficking, both inside and across national borders, is becoming increasingly sophisticated; the modern plague of synthetic opioids, arguably the first drug epidemic of the e-commerce era, is ravaging communities across the U.S.; and in the short 10 years since the inception of bitcoin, cryptocurrency presents a fertile financial environment to support illicit activities. Therefore, monitoring networked illicit activities in an era of hyperconnectivity and massive data will always be a pressing need. Network analysis capabilities are extremely helpful in advancing our ability to comprehend the operational structure of illicit networks and devise optimal disruption strategies. Natural language processing and deep learning remain key priorities (whether they are applied to online advertisements or trade-based money laundering), as it becomes increasingly impossible for individual investigators to navigate through massive amounts of data and identify hidden patterns without them. The CINA Science Committee provides this view of emerging opportunities and challenges in our respective fields, grouped here as priority issues which may shape criminal investigations in the near future.
Artificial Intelligence (AI) Applications
- Accurate face recognition (and even long-distance recognition of individual faces in crowds) is already being deployed for commercial and government purposes in some countries. Research on automatically annotating images and videos holds increasing promise for accurate text-driven search.
- AI applications in Natural Language Processing have strong transformative potential for investigations. This includes entity detection (finding people, companies, places, etc.), topic detection and tracking (following social media and news commentary about an evolving situation), and sentiment detection (determining opinions about topics). Breakthroughs are bolstered by the collection of background knowledge (gathered from the web or other sources) and automated inference (deduction to draw conclusions using learned rules).
- Knowledge graph techniques allow us to make information explicit, building networks of actors, interactions, and related elements from structured and unstructured sources, revealing the nature and full scope of illicit operations. Research on trend detection and causality discovery continues unabated, also using techniques developed in (and producing results for) aspects of Economics, Biomedicine, and Physics.
- Securing the cyberspace relies on the use of AI and machine learning generally, but is increasingly supported by a complementary array of breakthroughs in Biometrics (to guarantee identity), computer hardware (to prevent the introduction of malware and spyware), security of the Internet of Things, computer networks (for anonymity and censorship resistance), software (for deepfake detection), and blockchain technology.
Digital Forensics
- Data Access challenges include the need to extract data from an ever-growing ecosystem of devices, to include Internet-ofThings, embedded systems, vehicles, wearables, and many more. Investigators’ inability to access encrypted data poses another data access challenge. The final data access challenge is driven by the increasingly distributed storage of data, to include cloud storage in various jurisdictions and the opportunity for data to move between jurisdictions. The cloud also creates technical challenges, as cloud infrastructures are, by design, shared across multiple users and organizations.
- Data analysis challenges are driven by the immense and increasing volume of data available. Specific opportunities exist in automation of the triage and analysis processes so that human investigators can focus on leads, relationships, and content most likely to be of value to an investigation
- Data integrity challenges emerge from the development and availability of tools and techniques to manipulate digital content, e.g., images, videos (deep fakes), and audio (voice generators)
- Tools such as The Onion Router (TOR) enable mutually anonymous network connections, private VPNs provide onesided network anonymity, and compromised systems in the form of bots or live relay points obfuscate attribution to the true source of network traffic. Techniques are occasionally developed to trace such connections, but these techniques are often exposed in court proceedings which renders the techniques much less effective against an observant and dynamic cybercriminal community.
Forensic Science and Interaction with Victims, Suspects and Detainees
- Rapid DNA screening technology at the border can counter human trafficking and assist with a broad range of applications (e.g. asylum cases, refugee reunification, and overseas adoption cases), but there are still legal and policy challenges to overcome regarding its use.
- More efficient methods are needed to support opioid detection in shipments entering the country, as well as for chemical detection screening to disrupt the flow of opioids into the United States.
- A better understanding of behavioral analysis concepts, ranging from pre-offense, offense and post-offense behaviors, as well as training in interview concepts would benefit investigations. This would advance our ability to detect suspicious behaviors at the border which may be associated with the illegal transportation of people or goods.
- We need to advance our understanding of, and our ability to detect, patterns of warning behaviors by offenders in the period leading to mass shootings or other acts of terrorism.
- There is an increased focus on immigration issues, including interviewing people at the border and interviewing human trafficking victims. The increased focus on obtaining (more) information from suspects/detainees, as opposed to obtaining confessions, puts a renewed focus on the development of robust interview techniques to obtain more and accurate information.
Strategic Interdiction
- There is a developing interest in predictive analytics, to help resource allocation from local to federal scales.
- The Law of Crime Concentration suggests that a small part of a city will host the majority of crime-related problems. This seems likely true of such problems as illegal border entry and related ICE problems. Further research on hot spot analysis for problems at the level and geography of ICE concerns will be quite beneficial.
- There has been a growing recognition of the importance of diffusing programs and practices that have proven to be effective. This has led to the development of rigorous program evaluations in fields from health to policing. DHS operations can benefit from assessments of the effectiveness of DHS investigations.
- There is a growing awareness of the importance of assessments of the community impact of investigations and interventions. This will advance our abilities to proactively identify optimal intervention strategies, and also to better articulate to local communities how such investigations improve the everyday lives of the general public.