Data and analytics: Powerful weapons in the fight against human trafficking and modern slavery

14 September, 2023

Authors:
Liz Warner, Director, Financial Crime, PwC Canada
Pallavi Nautiyal, Manager, Financial Crime, PwC Canada

The recent enactment of the Fighting Against Forced Labour and Child Labour in Supply Chains Act, 2023 (Modern Slavery Act) has brought forward significant reporting obligations for a wide range of Canadian businesses and importers as well as government entities. With this topic top of mind, financial institutions may be wondering what they can do to help detect and prevent human trafficking and modern slavery.

Human trafficking includes recruitment or abduction of victims through deception or force (e.g. kidnapping, false marriage, offers of employment) and harbouring or transporting victims where they’re exploited or sold to other traffickers. Modern slavery includes exploiting victims for debt bondage, sexual exploitation, organ trade or forced labour. The victims may be manipulated by force, through withholding identity documents, illegal arrangements to work “off” recruitment fees or other forms of coercion.

Many typically think of money laundering predicate offences, such as drug trafficking or the proceeds of fraud, but there are other serious crimes, such as human trafficking and modern slavery, happening across Canada. These crimes can be detected through anti-money laundering and anti-terrorist financing controls and systems that financial institutions already have in place.

Where to start: Detecting human trafficking

Several years ago, a public-private partnership between the Financial Transactions and Reports Analysis Centre of Canada (FINTRAC) and Canadian banks was formed to detect human trafficking by focusing on the money laundering aspect of the crime. This initiative identified several indicators to better inform the detection of trafficking through the Canadian financial services industry. In less than two years after the inception of this partnership, 270 disclosures relating to 534 subjects were issued, of which 90% were previously unknown to law enforcement. The most common form of human trafficking activities is forced labour related to sex work or sexual abuse.

To track cases of human trafficking, it’s essential that, during the transaction monitoring and investigation processes, employees aren’t just checking for money laundering/terrorist financing activities, but also specifically for human trafficking activities. Investigation notes and internal/regulatory reports should highlight suspected potential connections and concerns relating to human trafficking activity

Key indicators of human trafficking and modern slavery activity

While many of the known victims of human trafficking are females under 25 years old,² trafficking occurs across gender, nationality and age. A study conducted by the United Nations University found that individuals from certain countries (e.g. Nigeria, Philippines) and industries (e.g. personal services industry) are particularly vulnerable to human trafficking and modern slavery activities.³ A combination of these two factors may be used by organizations to designate customers as high risk for human trafficking and subject them to enhanced monitoring.

Red flags have been documented for both victims and traffickers,⁴ with some examples below:

Indicators for the victim:

Large number of taxi/transportation, pharmacy or hotel payments
ATM transactions in cities across the country
Transfer from a third party to one person

Females under 25 years old

Indicators for the trafficker:

Multiple email addresses, phone numbers and/or physical addresses
Email address found on escort websites
Cash deposits in cities where the customer doesn’t reside or conduct business
Frequent payments to online escort, adult content, advertising services, online classifieds or website hosting companies
Frequent sending and/or receiving of funds via cryptocurrency
Lavish lifestyle purchases
Frequent purchases and/or use of prepaid cards
Peer-to-peer payments followed by withdrawal or transfer
Frequent cash withdrawals from white-label ATMs
Frequent transaction cancellations
Transactions that often happen outside of normal business hours
Car rental within city of residence
Frequent low value and/or abnormally high fast-food spending
Multiple cell phone and internet bills on the same account
Rent, property tax and utility bill payments in different locations in the same month
Owner, operator or employee of spa, massage parlour, strip club, motel, real estate rentals and/or property management
Males aged 24–36 years old or females aged 27–32 usually linked to an involved male

Now that we know the indicators and customer features that may be considered high risk for human trafficking activities, how do we look for this information in the billions of transactions, accounts and customer records housed in the databases of Canadian financial institutions?

Leveraging tools to detect patterns of behaviour

From a data and analytics perspective, there are clear challenges that have emerged as a result of trying to identify human trafficking activities. Most notably, there can be many similarities between a transaction indicative of human trafficking activity and a legitimate transaction. For example, someone could regularly withdraw cash at an ATM late at night for legitimate reasons. All activities must be considered in context and in addition to other transaction or customer profile indicators to determine if the activity is indeed suspicious.

Using machine learning and artificial intelligence (AI) to detect patterns of behaviour that traditional predictive models and/or analytics wouldn’t have previously been able to identify offers a way to detect complex combinations of activity that may be indicative of human trafficking. And the accuracy of these models automatically improves over time. Machine learning can be used to segment customers and detect atypical spending patterns compared to peer groups. This information can then be layered with additional human trafficking indicators, such as where and when the transactions are taking place, specific customer attributes like age and occupation and linkage to external data sources, such as escort websites, to further refine the model.

Graph algorithms and databases are also powerful tools to detect hidden relationships between shared personal attributes, complex payment chains, country risk concentration, funnel accounts and/or rapid fund movement activity. These can be especially useful in bringing together disparate data to identify previously undetected relationships.

Why this matters to financial services

At PwC Canada, we help our clients build trust in society by identifying, monitoring, assessing and reporting on human trafficking and modern slavery risks. We help them develop, tune and implement analytics and machine learning models. And we help them rethink the way they look at financial crime by bringing in smarter and more effective ways to not only be compliant, but to do something meaningful and impactful at the same time.

How can your organization help prevent human trafficking and modern slavery?

All organizations have a role to play in preventing human trafficking and modern slavery. While the new Canadian legislation represents obligations for many organizations, the reputational risk and ethical considerations are just as important for many others.

Financial institutions play a crucial role in detecting and preventing these crimes. We work with our financial services clients to enhance and adapt their anti-money laundering processes and systems to include the flags and indicators for human trafficking.

For other types of organizations, it will be key to understand modern slavery, identify risks in the supply chain and implement policies, processes and monitoring to detect and avoid forced labour to comply with the new Canadian legislation.

At PwC Canada, creating trust in what matters is core to our values and purpose. If you’re interested in learning more about how to prevent human trafficking and modern slavery, our community of solvers would love to hear from you.

_{1. Project PROTECT: Public Service Renewal in Action, Financial Transactions and Reports Analysis Centre of Canada (FINTRAC), accessed August 23, 2023, https://fintrac-canafe.canada.ca/emplo/project-projet/psr-eng.pdf.}

_{2. Updated Indicators: Laundering of proceeds from human trafficking for sexual exploitation, FINTRAC, July 2021, https://fintrac-canafe.canada.ca/intel/operation/oai-hts-2021-eng.}

_{3. Detecting Financial Flows of Modern Slavery and Human Trafficking: A Guide to Automated Transaction Monitoring, Dr. Frank Haberstroh and Simon Zaugg, United Nations University: Centre for Policy Research, May 2023, http://collections.unu.edu/eserv/UNU:9113/Indicators_Guide_Final.pdf.}

_{4. FATF Report: Financial Flows from Human Trafficking, FATF - APG, July 2018, https://www.fatf-gafi.org/content/dam/fatf-gafi/reports/Human-Trafficking-2018.pdf.}