A third company linked to suspected money laundering

AI system

System: Bank F network (graph) analytics solution implemented in Hong Kong in 2013
Model: Not assessed
Release: Not assessed
Application developer: Bank F (anonymized subsidiary and regional head office of a large global bank); the report names no vendor
Role: Not assessed
Configuration: Not assessed
Inputs and outputs: Not assessed
Autonomy: Not assessed
Workflow: Integration of wholesale client, correspondent banking, trade, payments and external registry data in one network model. A human intelligence lead (the relationship manager's TBML suspicion) used as the seed for network expansion.
Human oversight: Not assessed
Relation to AI: The HKMA report attributes the identification of Company C to the network analytics tool's matching of linked phone numbers and email addresses. A human relationship manager supplied the initial lead. Whether the tool uses machine learning is not stated. The bank built and deployed the tool itself. The HKMA, which wrote the report, is the regulator.
AI contribution: The HKMA report attributes the identification of Company C to the network analytics tool's matching of linked phone numbers and email addresses. A human relationship manager supplied the initial lead. Whether the tool uses machine learning is not stated. The bank built and deployed the tool itself. The HKMA, which wrote the report, is the regulator.

Outcome identifiers: Not assessed
Benefit: even if 10% were unrecoverable the tool contributed around USD 50 million in savings
Category: Reduced Risk
Status: Not assessed
Beneficiary groups: Organizations. Society.
Beneficiaries: Not assessed
Reach: Not assessed
Distribution and access: Not assessed
Duration: Not assessed
Expected?: Yes
Costs and harms: Not assessed

Studies: Not assessed
Study design: Not assessed
Measurement: Not assessed
Setting and sample: Not assessed
Measures: even if 10% were unrecoverable the tool contributed around USD 50 million in savings
Comparison: Not assessed
Results: even if 10% were unrecoverable the tool contributed around USD 50 million in savings
Additional value from AI: Not assessed
Limitations: Not assessed
Denominators: Not assessed
Materials for replication: Not assessed

Participants: Developer: Bank F (anonymized subsidiary and regional head office of a large global bank); the report names no vendor. Operator: Bank F, Hong Kong.
Jurisdictions: Benefit: Hong Kong, China.
Ownership and control: Not assessed
Dependencies: Not assessed
Cross-border connections: Not assessed
Geopolitical implications: Not assessed

Conditions for replication: Data of sufficient quality across sources, which the report says Bank F stressed. Governance that lets analytic findings trigger account reviews and STR filings.
Replication attempts: 3 follow-up entries. Benefit recurred: 2. Outcome unresolved: 1.
Later findings: Nothing dated after the January 2021 HKMA report was found for Bank F's tool; the report's own scale-up figures predate it.
Availability: Not assessed
Institutional decision: Not assessed
Grounds: Not assessed
Action and result: Not assessed
Success criterion: Not assessed

Source numbers: AIBD-0174-source-0001–0002
Citations: Numbered references below.
Language: Not assessed
Source relationships: HKMA and Deloitte describe Bank F’s experience. The bank supplied the account. The USD 50 million figure assumes that 10% of the unsecured exposure would have been lost; it is not a measured saving. [1] [2]
Original evidence: HKMA and Deloitte describe Bank F’s experience. The bank supplied the account. The USD 50 million figure assumes that 10% of the unsecured exposure would have been lost; it is not a measured saving. [1] [2]
Access and versions: Not assessed
Supported findings: Not assessed
Evidence links: Not assessed
Evidence strength: Not assessed
Missing evidence: Not assessed

Assessed benefits: Organizations. Society.
Depth: Organizations: 2. Society: Not assessed.
Reasons: Organizations: The report records reviews of around 100 customer relationships. It also records a third linked company whose review revealed potential money laundering concerns. Losses avoided are estimated, not observed. Society: The foreword's statement about neutralizing mule account networks is general and not tied to a documented outcome.
Observed or intended: Not assessed
Rubric: Not assessed
Assessor and date: Not assessed
Alternative assessments: Not assessed
Fill: Independently reported

Kind: Not assessed
Related entries: Not assessed
Field: Anti-money laundering. Trade finance and trade-based money laundering.
Purpose: Identify high-risk customers, transactions and hidden relationships for AML/CFT and trade-based money laundering investigations.
Dates: Date basis unassessed: 2021 (precision unassessed). Updated: 25 October 2026.
Stage: Not assessed
Version: SHA-256: f71e252dd0e9…