Phoenix Compliance Essentials #18
Today’s focus: KYC and Artificial Intelligence
AI is reshaping operational practices across many sectors.
AML/CFT-P-C frameworks are no exception.
This transformation is two-sided: AI enhances KYC processes but also makes them easier to circumvent.
Deepfakes and forged documents are becoming tools for gaining access to business relationships.
➡️ How do deepfakes threaten KYC frameworks?
Deepfake fraud consists of impersonating an individual using AI-generated content capable of realistically replicating a person’s face, voice, gestures and speech patterns in order to deceive identity verification systems.
Deepfakes now represent a major operational risk for KYC frameworks.
Supervisory authorities are well aware of this threat: the FATF has published a report on these vulnerabilities and calls on institutions to strengthen their detection and control frameworks.
➡️ Document forgery: a threat amplified by AI
According to an Ipsos survey:
– Nearly 20% of professionals report having already encountered a forged document;
– This figure rises to 30% among senior executives.
These figures reflect a daily reality for businesses: document forgery has become a common and increasingly accessible threat, amplified by the rapid development of AI.
AI makes it possible to produce highly credible forged documents quickly, even by fraudsters with little or no specialist expertise.
As a result, in a KYC context, it is no longer sufficient to assess only the visual consistency of a document; its reliability and plausibility must also be assessed in light of the customer’s profile.
➡️ Securing your KYC framework against these new threats:
In practice, several good practices can be implemented:
– Train staff to detect anomalies.
– Adapt the level of verification to the customer’s risk profile.
– Cross-check the information provided against independent sources.
– Verify the authenticity of documents using official sources such as PRADO, which enables users to identify the expected security features of official documents.
– Verify the validity of identity documents through the MRZ (Machine Readable Zone).
– Use NFC technology to verify the authenticity of biometric passports and electronic identity cards by accessing the secure data stored on the chip.
– Use tools incorporating OCR (Optical Character Recognition) technology to extract data from documents (identity documents, supporting documents and legal documents) and automatically reconcile it with declared information or data obtained from third-party databases.
– Implement facial recognition and biometric verification tools.
