Ant Group's Full‑Graph Risk Control Architecture and Its Application in Combating Complex Fraud

The article presents Ant Group's full‑graph risk control system, detailing emerging fraud trends, the need for graph‑based anti‑fraud infrastructure, and the multi‑layer architecture that combines data cleaning, graph modeling, multi‑modal computation, and real‑time detection to tackle sophisticated, organized financial crimes.

AntTech
AntTech
AntTech
Ant Group's Full‑Graph Risk Control Architecture and Its Application in Combating Complex Fraud

At the recent Digital China Summit, Wang Xingchi, head of Ant Group's full‑graph risk control technology, publicly shared the architecture of Ant's intelligent risk control system "IMAGE", which includes interactive proactive risk control, edge‑cloud collaborative risk control, multi‑party risk control, and intelligent adversarial defense.

Graph technology is becoming a focal point in the risk control market because it can construct a risk relationship network, providing a holistic, relational view of risk chains and addressing the fragmentation of traditional risk control.

The presentation identified three upgraded risk trends: (1) the emergence of score‑running platforms and "water‑room" accounts that complicate fraud chains; (2) the rise of organized, team‑based fraud increasing collective risk; and (3) accelerated fund flow timing that makes single‑transaction analysis insufficient, especially in money‑laundering scenarios.

To meet these challenges, Ant built a next‑generation risk control infrastructure based on a full‑graph approach, which must handle both spatial (entity relationships beyond individuals) and temporal dimensions with millisecond‑level latency.

The full‑graph system integrates graph‑based gang detection, fund‑flow tracing for anti‑money‑laundering, and trusted‑network identification, supporting the entire risk lifecycle—from perception and identification to control, adjudication, and analysis.

Technically, the architecture consists of a data‑layer that normalizes heterogeneous, semi‑structured, and unstructured sources into a unified risk graph, and a computation layer that unifies graph query, mining, and learning via a multi‑modal DSL, allowing flexible triggering, materialization, and optimization for real‑time SLA requirements.

In production, Ant's risk graph has improved detection capability by 9.4× and case‑analysis efficiency by 90%, positioning the full‑graph as the foundational infrastructure for combating complex, cross‑domain financial fraud.

Original Source

Signed-in readers can open the original source through BestHub's protected redirect.

Sign in to view source
Republication Notice

This article has been distilled and summarized from source material, then republished for learning and reference. If you believe it infringes your rights, please contactadmin@besthub.devand we will review it promptly.

fraud detectionInformation Securitygraph technologyanti‑money laundering
AntTech
Written by

AntTech

Technology is the core driver of Ant's future creation.

0 followers
Reader feedback

How this landed with the community

Sign in to like

Rate this article

Was this worth your time?

Sign in to rate
Discussion

0 Comments

Thoughtful readers leave field notes, pushback, and hard-won operational detail here.