LGT (Liechtenstein Global Trust)
Benefits of Our Transformation
Implementing machine learning-based solutions to reduce false positives and enhance screening system performance for LGT Private Banking.
About LGT
LGT Bank is the largest private banking and asset-management group owned by the Princely House of Liechtenstein. With a global client base and strict regulatory oversight, the bank operates under complex financial-crime compliance obligations. Over time, outsourcing parts of its screening process to third-party providers had created inefficiencies and high false-positive rates, driving operational strain and increased costs. LGT sought to regain control, improve accuracy, and make its compliance systems more efficient without compromising regulatory integrity.
Results That
Matter
50%
Reduction in Hits
28.5% to 22.1%
Sanctions False Positive Rate
- 11%
False Positives Reduction
Streamlined monitoring workflows
Compliance Accuracy
The Opportunity
The bank recognised the opportunity to enhance its customer and transaction-screening systems to balance accuracy and efficiency. High false-positive rates were consuming valuable analyst time and inflating operational expenses. LGT wanted to reduce unnecessary alerts, streamline reviews, and improve the detection of true matches through smarter analytics and system optimisation. The challenge lay in fine-tuning thresholds and introducing advanced automation while maintaining strict risk-management standards.
Solutions Delivered
Beyond partnered with LGT to enhance performance through data-led optimisation and automation.
- Capability Assessment and Benchmarking: Identified root inefficiencies in the screening system, providing a data-backed roadmap for improvement.
Threshold Recalibration: Adjusted sensitivity parameters to minimise false positives while preserving risk coverage.
Machine Learning Post-Processor: Introduced an AI-based layer that further refined results, filtering out non-relevant matches and reducing manual workload.
Results & Impact
The programme delivered transformative results. LGT achieved a 50% reduction in the number of hits returned by the system, while the false-positive rate for sanctions screening fell from 28.5% to 22.1%. Following the implementation of machine learning enhancements, a further 11 percentage-point reduction was realised. These changes cut costs, improved operational efficiency, and freed compliance teams to focus on genuine risk cases. System accuracy and regulatory assurance were both strengthened, delivering a more balanced and sustainable compliance model.
Lesson Learned
LGT’s journey demonstrated that operational efficiency in financial crime screening comes not from wholesale system replacement but from targeted, data-led optimisation. Through precision thresholding and machine learning, the bank reduced noise and increased reliability — allowing its teams to focus on true risk and strategic compliance tasks. The experience proved that innovation in regulatory operations requires a measured blend of data science and governance discipline to achieve lasting impact.
