In the previous BusinessForensics article, we discovered that approximately 61% of financial service professionals find it difficult to spot signs of human trafficking. Therefore, aside from explaining what human trafficking looks like, and how to identify various red-flags, it’s also important to elaborate on which approach and technologies one should use to combat human trafficking.
How to start detecting human trafficking
In order to monitor and detect financial crime, as a whole, and human trafficking, specifically, using the indicators and red-flags mentioned in the first article, we need data! But, there is so much data that it is almost undoable, simply, to start monitoring. In order to do it right, we need to start at the end.
We need to specify the correct questions that will help us gather the correct data.
- What happened?
- Who is involved?
- ¿Cuándo ocurrió?
- ¿Dónde ocurrió?
- ¿Con qué medios o métodos?
- Why did it happen?
These are all questions that an investigator will ask themselves when investigating this type of crime, or any financial crime. Answer these questions using critical thinking, where you focus on the facts that help determine which data to use.
What data do you need for investigative case management?
Fighting crime, in contemporary times, means utilizing a data-driven approach. Not only in monitoring and detection, but also during analysis and investigation. You need to visualize what you found to substantiate your story. In the example, you see a transaction where money was transferred from account A to account B. Now, let’s assume this transaction is executed by a human trafficking organization.
In order to extract the correct data points from the transaction, we should use the questions we formulated.
- Who is involved? Account name: “Mr. X”
- Where did it happen? Customer Country: Romania
- Where did it happen? IP Address: 192.168.1.1
- With what means or methods? Means of payment: pre-paid card
Every transaction can hold up to hundreds of data points, usable for monitoring and detection. The same goes for events, or contextual information like identification documents, address information of the client, phone numbers, and email addresses specified by the client. In order to gather the correct data, one needs to know what needs to specify the required end-result.
What technology should you use?
If we leave out the possibilities your organization has, or offers, in terms of budget, people, and amounts of data, then we can focus on the ultimate situation in terms of technology.
If we look at the technology components that are needed, we can divide them into two separate modules: i) monitoring and detection, ii) investigative case management or incident register.
Monitoring and detection
When fighting crimes like human trafficking, transactions — and, in this case, transaction monitoring — is a good place to start. The technology used should be flexible enough to incorporate modus operandi, when needed, and smart enough to identify new modus operandi. This makes me believe that you should use a transaction monitoring system that can run scenarios with static rules, but is also capable of running monitoring based on profiles and various models created for supervised machine learning.
I do believe in the power of unsupervised machine learning, but not yet in the case of fighting financial crime — especially since unsupervised machine learning is a great new asset, but it can learn the wrong things, too.
So, I encourage the use of supervised machine learning where the data used is labelled. This technology will really help you in the process, but the human stays in the loop.
Investigative case management
Good case management, or incident management, as it is often called, has several functionalities that are absolutely necessary.
- (Smart) Workflow: to configure and use the processes and policies
- Audit trail: to document and track the actions that one took during the investigation
- Attachments: to store the relevant data, such as contracts, emails, etc.
- Visualization: to present the investigator with actionable intelligence as a necessary layer on top of situational awareness
Next to these basic functionalities, a good system should provide the organization with the ability to work together over different departments, and even subsidiaries in different countries, using the same data. This way, the knowledge is shared and the quality of the investigations improves.
Sharing knowledge is something I truly believe in, and I think initiatives like “Transaction Monitoring Nederland (TMNL),” Data Sharing Coalition EU, the ECCFI, and NGOs, like Sustainable Rescue, are helping to bring us forward in fighting human trafficking and financial crime.