How Standard Bank Group automates AI risk assessment for generative AI
Machine Learning Blog
This article describes how Standard Bank Group automated AI risk assessment for generative AI applications using AWS services, transforming governance from a manual bottleneck into a scalable, continuous process.
- Pre-assessment agent scans code repositories and documentation to populate a 40-question Well-Architected Tool lens with evidence
- Pipeline examines system documentation, policies, source code, and live application calls against international AI standards and internal policies
- Foundation model on Amazon Bedrock acts as judge using LLM-as-a-judge technique to evaluate evidence against 300+ controls across eight responsible AI dimensions
- Assessment produces executive presentations, mitigation checklists, technical reports, and CloudWatch dashboards for different audiences
- Reduced traditional two-week manual review to two-hour automated report; delivered 14% improvement in use-case risk scores after remediation
- Integrates into use case lifecycle at proof-of-concept, build, and production stages with continuous reassessment after launch
By treating policies as executable tests rather than documents, Standard Bank Group enables faster AI innovation while maintaining governance rigor and regulatory auditability.
The AWS News Feed is currently looking for gold sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.
Related articles
2026
2026
2024
2026
The AWS News Feed is currently looking for silver sponsors. If you want to support the AWS community and reach a large audience of AWS professionals, consider sponsoring the AWS News Feed.