Case Study: Commonwealth Bank of Australia cuts scam losses with real-time GenAI from H2O.ai
Key results
The challenge
Commonwealth Bank of Australia (CBA), the country's largest bank serving more than 16 million customers, faced growing customer scam and fraud losses that conventional detection methods struggled to contain. The bank needed real-time capabilities to keep customers safe across fraud detection, anti-money laundering, and customer service.
The solution
CBA partnered with H2O.ai to deploy real-time generative and predictive AI for fraud and scam prevention. The bank trained more than 900 analysts on H2O.ai and applied the models across use cases including real-time fraud detection, anti-money laundering, and transaction abuse monitoring.
“From our peak in 2022, we've reduced customer scam losses by 76%. AI has allowed us to keep our customers safe in ways that simply weren't possible before.”
DJDan JermynChief Decision Scientist, Commonwealth Bank of Australia
The results, in context
CBA reduced customer scam losses by 76% from its 2022 peak. The bank reports a 70% scam reduction and a 30% fraud reduction, with 900+ analysts trained on H2O.ai across a base of more than 16 million customers.