Neo4j Releases GraphAware Financial Crime Intelligence for Banks and Insurers
Neo4j launched a graph-based financial crime product for banks and insurers, its first release since buying GraphAware in August, as Interpol figures put global fraud losses at $442 billion.
The new product directs at private-sector financial institutions the same graph-based approach Neo4j pitched to government agencies in June, when it announced the GraphAware deal and positioned GraphAware's Hume software as an alternative to Palantir Technologies Inc.'s Gotham.
Neo4j cited Interpol data to frame the market. An Interpol threat assessment published in March estimated that financial fraud cost victims $442 billion globally in 2025. A single Interpol operation this year led to 5,811 arrests across 97 countries and territories and $293 million in intercepted funds, the agency said in July. Neo4j said regulators are shifting more of the burden for prevention onto financial institutions, often with expensive fines attached, while criminals use artificial intelligence to scale up their operations.
The software joins data held in separate systems into a single graph that analysts can query to trace links across accounts, transactions and devices. Neo4j describes that kind of multihop reasoning as a core strength of graph databases.
Underneath the product sits a reusable knowledge layer that Neo4j also markets as grounding for enterprise AI. Each case worked adds to it, the company said, so that later monitoring can draw on what earlier investigations turned up.
The workflow is organized into four stages. Signal covers detection and looks for suspicious patterns buried in connected data. Alert then passes investigators deduplicated warnings along with the context behind each flag. In the Investigate stage, graph analytics trace linked records, and case-specific data from third parties can be pulled in where internal records run out. The final stage, Decide, logs the outcome of each case and preserves the relationships and provenance behind the call as longer-term evidence.
"Every fraud involves a network," said Michael Down, global head of financial solutions at Neo4j. Storing relationships natively lets a graph platform hop between data points quickly enough to surface suspicious behavior, he said.
Neo4j said its software already supports fraud detection or compliance work at large institutions including BNP Paribas SA, UBS Group AG and Zurich Insurance Group AG. Klarna Group plc is among several fintechs using it for AI projects. Across all industries, 84 of the Fortune 100 run Neo4j, the company said in June.
SiliconANGLE Media's livestreaming studio, theCUBE, will cover Neo4j's GraphSummit on Sept. 24, where the company is expected to argue for graph technology as the knowledge layer for enterprise AI.