UBS
3 documented AI deployments at UBS, across 1 industry. Each one links to the public source it was taken from.
Public AI record
UBS has 3 documented public AI cases on the hub, too few for a score. Customer service automation and fraud detection are common among peers but not yet documented at UBS.
A record score appears once UBS has 5 documented public cases. The peer comparison below already uses what is on record.
First documented case Apr 2025, latest May 2025. No new deployment documented since May 2025. Cloud providers named: Microsoft.
Compared with peers
Against 9 other banking companies on the hub.
What peers document
- Customer service automation6 of 9 peers67%
- AI development platform3 of 9 peers · here too33%
- Fraud detection3 of 9 peers33%
- Risk assessment3 of 9 peers · here too33%
Gaps against the industry
- Customer service automation: 67% of peers, not documented at UBS yet
- Fraud detection: 33% of peers, not documented at UBS yet
Signature use cases
No peer on the hub documents semantic search.
Stack and approach vs peers
- Peer cases name Microsoft 61%, AWS 18%, GCP 18%.
- AI agents: 33% of UBS's cases vs 24% for peers.
- Cases from the last 12 months: 0% vs 26% for peers.
Deployments
3 deployments at UBS, newest first.
- 1 May 2025FinanceSwitzerland
UBS RiskLab AICE: governed GenAI platform for data scientists
UBS · AI development platform · Microsoft
UBS RiskLab's AI Common Ecosystem (AICE) is a governed generative AI platform on Azure for more than 1,200 UBS data scientists. It provides governed production data access, managed LLM endpoints, tenanted RAG environments and AI agent deployment pipelines. Its VEGA vector store runs on Azure Database for PostgreSQL. RiskLab reports that taking an ML model into production dropped from six months to one month, and provisioning a vector store from days to less than 30 seconds. The source is a post by UBS RiskLab on Microsoft's Tech Community blog.
Production
↓ 30 stime - 27 Apr 2025FinanceSwitzerland
UBS Leverages Microsoft Azure AI for Fraud Detection
UBS · Risk assessment · Microsoft
UBS has dramatically bolstered its fraud detection and operational processes using Microsoft Azure AI. With the implementation of 'UBS Red', a custom AI assistant based on Azure AI solutions and Azure OpenAI Service, UBS enhanced client advisory efficiency. Over 60,000 investment documents were digitized into a searchable format. They established an AI Hub and data mesh framework to expand AI technologies organization-wide.
AgenticProduction
- FinanceSwitzerland
UBS Legal AI Assistant (LAIA): semantic search over legal documents
UBS · Semantic search · Microsoft
UBS's legal team built the Legal AI Assistant (LAIA) with Azure OpenAI Service and Azure AI Search to search a library of about 26 million legal documents in multiple languages by meaning rather than keyword. According to Microsoft, UBS legal experts can now run complex queries within minutes, not hours, and UBS is expanding the tool globally.
Production
What it builds
AI development platform is the most common pattern in these deployments: 1.
Cloud providers named
Microsoft is named in 3 of these deployments.
- Microsoft100% of these3