Published 10 Jun 2024· via HSBC· News article
HSBC revolutionizes crime detection with AI
HSBC · Finance · United Kingdom
The source does not publish a single comparable figure. Reported results are listed below.
Summary
HSBC, in collaboration with Google, employs advanced AI systems to prevent financial crimes more effectively. Their 'Dynamic Risk Assessment' AI processed 1.35 billion transactions monthly across 40 million accounts, significantly enhancing detection accuracy. HSBC's AI-driven capabilities reduced false-positive alerts by 60% while increasing detection rates for financial crimes.
The problem
- 01Detect financial crimes effectively at immense transaction volumes.
- 02Manage false-positive alerts that hinder operational efficiency.
- 03Ensure customer experience remains unaffected during transactions.
What they built
- 01Developed the Dynamic Risk Assessment AI system in partnership with Google.
- 02Employed advanced algorithms to identify financial crime patterns.
- 03Processed over 1.35 billion transactions monthly across 40 million accounts.
- 04Integrated the system seamlessly with global banking operations.
Reported results
- 01Enhanced crime detection accuracy, increasing incidents identified by 2-4 times.
- 02Reduced false positives by 60%, optimizing efficiency.
- 03Accelerated processing timelines from several weeks to days.
- 04Maintained operational integrity and high customer approval ratings.
- Improved financial crime detection
- Reduced false-positive alerts
- Accelerated processing timelines
- Maintained customer experience during transactions
Source
How this deployment compares
Computed from the library · data 10 Nov 2026HSBC has 4 deployment records in the library, from 2021 to 2024. It is one of 78 risk assessment deployments, 37 of them in Finance.
HSBC in the library
4 deployment records from HSBC, 2021–2024
- Mar 2021HSBC deploys Dialogflow to reduce policy-expert call burdenCustomer service automation
- Sept 2022HSBC boosts customer service automation with AI conversational botsCustomer support automation
- May 2024HSBC automates small business accounting with embedded AI-powered solutionCompliance automation
- Jun 2024This record · Risk assessment
Still referencedThe source is still online, but no newer HSBC record has appeared yet.
Same pattern, whole library
One of 78 risk assessment deployments
- Finance accounts for 37 of them (47%), more than any other industry.
- 18 records of this pattern came before it; 47 have been published since.
- Only 19 of the 78 put a number on the result (24%).
- Microsoft68%
- GCP17%
- AWS13%
- Alibaba1%
- Azure OpenAI 21
- Azure AI 17
- Azure 11
- Amazon Bedrock 9
- BigQuery 7
- Microsoft Fabric 5
Outlined: used in this record.
Shares describe this library, not the whole market.
Results against peers
What peers report
- Time & speed36 figures, Finance · middle half 39%–75%57%median reduction
- Risk, reliability & safety13 figures, all industries · middle half 60%–90%70%median reduction
- Customer experience50 figures, all industries · middle half 20%–90%28%median increase
Peer figures are as reported by their sources; percentage claims only.
How we checked this record
Evidence tier: supported
A credible source with a named customer, but not the organization's own account.
- Named organization
- HSBC
- Who published the source
- A third party (press or analyst)
- Deployment status
- Production (confidence 95%)
- Quantified result
- The source reports at least one figure
- Source link
- Online at last check
Last re-checked 13 Jun 2026.
This record is evidence for
Same pattern, other organizations
All risk assessment deployments- 27 Jul 2026Tech & CommsHong Kong
Trusta Labs — Web3 sybil attack prevention and knowledge-graph enhanced profiling using Alibaba Cloud data/AI services
Trusta Labs · Risk assessment · Alibaba
Trusta Labs is a Web3 infrastructure company in Hong Kong that builds security and data infrastructure through AI. It uses Alibaba Cloud elastic computing, storage, data management, and analytics services to match on-chain assets with the right users and to support AI-powered Sybil attack prevention, knowledge graph-enhanced profiling, and user targeting.
- 15 Jul 2026InsuranceUnited States
Delos Insurance Solutions uses Google Earth Engine and Vertex AI to underwrite wildfire insurance
Delos Insurance Solutions · Risk assessment · GCP
Delos Insurance Solutions needed to make wildfire-risk underwriting viable for California homeowners who were often considered uninsurable by other carriers. The company built wildfire-behavior and geospatial risk models with Google Earth Engine, then used Colab notebooks and Vertex AI pipelines for model training and deployment, plus Google Maps Platform for a quote-to-buy underwriting portal. The solution let Delos price policies more accurately, offer coverage in high-risk areas, and support agents with a fast online underwriting workflow.
−80%data preparation time - 8 Jul 2026FinanceCanada
KOHO: Generative AI on Amazon Bedrock for credit underwriting and fraud detection
KOHO · Risk assessment · AWS
KOHO rebuilt credit infrastructure to assess credit-invisible customers using behavioral data such as rent, utility, and phone payment records. The company built Kortex AI on Amazon Bedrock to extract behavioral insight for underwriting and used AI screen detection to reduce account takeovers.
−70%account takeovers - 7 Jun 2026Public SectorUnited States
NYC Cyber Command builds SRE-backed threat-management/analytics platform on Google Cloud
New York City Cyber Command · Risk assessment · GCP
New York City Cyber Command (NYC3) built a security log aggregation platform on Google Cloud to keep city security systems available 24/7 and support risk alerts, visualization, and analytics for 100+ city agencies. The organization adapted Google’s Site Reliability Engineering model, created an internal SRE team, and trained/cross-trained engineers across development, networking, and operations to improve reliability and supportability. NYC3’s internally developed Juggernaut threat-management system parses real-time cybersecurity data from disparate sources and alerts the team when something seems amiss.