Key Takeaways from the ACPR Conference 2024:

1. Strengthening KYC (Know Your Customer) Requirements

Context and Challenges:

  • KYC requirements are essential for identifying customers and assessing the risks associated with their activities. However, with the rise of digital technologies, traditional KYC methods are reaching their limits in terms of efficiency and speed.
    The challenge lies in processing millions of data points in real-time while complying with regulations.

Issues Highlighted:

Increased Complexity of Customer Data:
  • Growth of multichannel interactions (mobile, digital, branches).
  • The proliferation of digital identities requires more sophisticated verifications.
High-Risk Customers:
  • Greater presence of politically exposed persons (PEPs) or complex structures like trusts.
  • Hidden transactions using proxies or shell entities.
Regulatory Pressure:
  • Compliance with European and international regulations, such as AMLD5/AMLD6 directives.

Proposed Detailed Solutions:

Digitization of Processes:
  • Automating customer identification through biometric tools (facial recognition, fingerprint scanning).
  • Leveraging interconnected databases, such as national and international business registries
Proactive Anomaly Management:
  • Detecting inconsistencies between customer-provided data and public records (e.g., addresses or names linked to high-risk zones).
  • Continuous monitoring of changes in customers’ risk profiles (dynamic surveillance).
Enhanced Due Diligence:
  • Obligation to document all interactions with PEPs or clients linked to sensitive jurisdictions.
  • Increased checks on beneficial owners of companies and complex ownership chains.

2. Artificial Intelligence and Big Data

Context and Role:

  • Traditional detection tools based on predefined rules are becoming obsolete due to the rapid evolution of money laundering techniques and data volumes.
  • AI and Big Data are critical for automating surveillance processes, detecting complex patterns, and reducing false positives.

Real-World Use Cases Discussed:

Improved Detection:
    • Predictive models to identify unusual activities based on historical data.
    • AI to uncover hidden relationships between seemingly unrelated entities.
    • Proactive detection of complex financial setups, such as layering (dissemination of illicit funds across multiple accounts and jurisdictions).
Reducing False Positives:
  • Traditional systems generate a high volume of unnecessary alerts, overburdening compliance teams.
  • AI algorithms analyze multiple variables (transaction type, time, location) to refine detection criteria and prioritize relevant alerts.
Handling Massive Data Volumes:
    • Leveraging structured data (bank transactions) and unstructured data (emails, documents, social networks).
    • Real-time analysis of millions of transactions and generation of synthetic reports for compliance teams.

3. Challenges of Using AI :

Regulatory Oversight :
    • Regulators demand greater transparency in the design and use of AI models to avoid biases.
    • Obligation to document and explain decisions made by algorithms.

Ethics and Bias:

    • Risk of excluding certain profiles unintentionally reinforcing discrimination.

Team Training :

    • Employees need to understand the results produced by AI tools and integrate them into their daily work.