The Commonwealth Bank of Australia (CBA) is integrating generative AI across its operations to improve customer service, simplify processes, and strengthen security. Speaking at the The Rising 2025, India’s biggest summit on women and tech in AI, Nidhi Sinha, general manager, chief data analytics office (CDAO) at CBA India, revealed that the bank built its generative AI chatbot, CommBank, in less than six weeks.
“A lot of organisations are experimenting with generative AI,” Sinha said, “But we are using it to lift and accelerate what we are currently doing.” She added that while the initial focus was improving internal operations, CBA has now extended these capabilities to customer-facing solutions.
One such initiative was launched in November for business banking. Business customers, who often have complex queries about products and payments, previously had to navigate through 80 different FAQ documents or contact call centres, resulting in delays.
“We have embedded the generative AI solution within the app,” Sinha explained. “It fetches data from all these documents, allowing customers to complete transactions without leaving the app.” The chatbot understands context and provides relevant answers, reducing the need for external searches or call centre support. “Best of all, this entire solution was developed in just six weeks, from infrastructure provisioning to deployment.”
She added that CBA built this solution on AWS, one of its strategic partners. “The pace of change has never been this fast, yet it will never be this slow again,” Sinha remarked, quoting an internal mantra that reflects the rapid advancements in AI adoption.
Notably, CommBank entered a five-year strategic collaboration with Amazon Web Services (AWS) earlier this year to continue as the bank’s preferred cloud provider.
The bank currently has over 60 generative AI use cases, with many already live for both customers and internal users. These solutions drive efficiency and improve customer interactions.
To support AI-driven transformation, CBA has established the GenAI Council, a leadership body that includes senior executives and the CEO and is focused on AI acceleration.
“A combination of oversight and a combination of federation is really helping us accelerate at a very, very fast pace,” she further said. “Different teams run with their own use cases, while central oversight ensures alignment and scalability.”
CBA’s investment in AI spans more than a decade. “We have been recognised as the number one bank in AI in the Asia-Pacific region for two consecutive years and globally as the best in responsible AI,” Sinha said. “We are also working with the Australian government to develop AI principles for the country.”
She further shared that over 60% of the population in Australia uses CBA in some form, with one-third considering it their primary financial institution.
The bank has prioritised responsible AI, integrating governance frameworks to ensure safe implementation. “For example, checks on groundedness prevent hallucinations, profanity filters maintain appropriate interactions, and jailbreaking safeguards ensure models are not misused. These controls are centrally managed and available for all teams.”
CBA is also accelerating its data strategy, moving from on-premise to cloud. “Initially, this was planned for 18 months, but by leveraging AI, we are completing it in nine months,” Nidhi said.
“By June, all our data will be on the cloud, providing practitioners with high-quality data to accelerate use cases.”
Collaboration is key to CBA’s AI journey. “We have world-class partnerships with AWS, Microsoft, and Anthropic, giving us access to cutting-edge AI capabilities and top talent,” she said. “These partnerships are critical in our generative AI journey.”
AI-Driven Financial Management
One of CBA’s AI-driven solutions focuses on financial management. The bank uses predictive analytics to help customers manage their finances effectively. “We have a product where we use AI to predict future cash flows and nudge our customers at the right time,” Sinha said. This proactive approach allows customers to plan better and avoid financial stress.
CBA also applies AI to disaster response. Given Australia’s exposure to natural disasters, the bank integrates external weather data with its customer information. “We use data and AI to identify customers and communities proactively who would be impacted by some calamity, and we reach out to them to provide our help,” she said.
Personalised Customer Engagement
CBA has been personalising customer interactions for over a decade. “In 2015, we launched something we call the Customer Engagement Engine,” Sinha shared. This AI-driven platform connects all customer interaction channels—mobile banking, branch visits, and call centres—to offer real-time recommendations.
“When a customer comes into any of these channels, there is something called next best conversation (NBC) that is surfaced to them,” she explained. “It could be about an offer, a service, or even a simple happy birthday message.”
She added that the system processes over 3.1 trillion data points and runs 2,000 adaptive models to ensure relevance.
This AI-driven approach has also influenced customer engagement through CBA’s loyalty program, Yello. “What we have seen in the last five years is that customers engaged with this program log into our app, on average, 67 times a month—twice a day.”
AI for Fraud Prevention
CBA has made significant strides in fraud prevention using AI. “With AI evolving, threat actors and scammers also have access to these technologies, and fraud is increasing globally,” Sinha warned. She cited global statistics showing a $1 trillion loss due to fraud last year.
However, CBA has successfully reduced scams by 70% in the past two years. “This is a big achievement, though we aim to do more,” she explained. The bank monitors 20 million transactions daily, detecting fraudulent activity in real time. “Within 10 milliseconds, an alert is sent to the customer, allowing them to take immediate action.”
CBA has also introduced features like NameCheck and CallerCheck to prevent mistaken or fraudulent transactions. “This has actually helped us save $650 million per customer,” she shared.
Preventing Payment Abuse
One of the more unique AI applications at CBA involves preventing abuse through payment messaging. “We discovered that some individuals were misusing the payment description field to send abusive messages,” Sinha said.
In response, CBA launched a profanity blocker in 2021. “At the time of payment, if abusive words are detected, the transaction is blocked immediately,” she explained. However, the challenge went beyond explicit words. “Even simple phrases like ‘I love you’ can be threatening in certain contexts,” she said. The bank has since developed AI models to detect harmful intent, ensuring payments are not used as a tool for harassment.
Talent Development
Another focus area is talent development. “People are the cornerstone of our AI initiatives,” Sinha said. “Five years ago, we set up CBA India, and today, we have 46% workforce diversity, with 41% representation in leadership roles…An inclusive workforce enables us to understand customers better and think beyond traditional banking.”
CBA’s AI journey is not just about technology but also about fostering a culture of experimentation, innovation, and responsibility. “We are creating a model where people raise their vision to do more with AI,” Sinha concluded.
