GenAI will transform banking with more focused customer experiences: Deutsche Bank CTO Leukert

Just as e-commerce transformed physical interactions, Generative Artificial Intelligence (GenAI) is set to revolutionise banking for customers.

"This advancement will drive unprecedented digitalisation in banking, creating a more tailored and efficient service than ever before," he said in an interview during a recent visit to Bengaluru.

The impact of AI, particularly GenAI, on various industries, especially banking, will be “profound and transformative," Leukert insists. But he adds that user experiences need to be streamlined into a unified approach. 

Leukert calls this the “Concept of One." He cites the example of Apple, which provides a single, consistent way for users to download and access a diversity of apps in the App Store—from weather forecasting to mobile banking. This uniformity, he explains, simplifies user interaction, despite the underlying complexity of the services offered.

"In the realm of AI-enabled services, the aim is to create a foundational level of service that can be adapted and improved across multiple use-cases," Leukert adds.

AI applications

For instance, AI can handle the processing of a variety of documents, whether for customers or internal use. This includes automating tasks related to mortgage applications or managing insolvency notifications, Leukert points out. AI's ability to understand and act upon diverse documents enhances operational efficiency and accuracy. 

Another application of AI is in software development, where it can make code more robust through auto-completion and error correction, he underscores. This speeds up coding by reducing syntax errors and suggesting completions, although human oversight remains necessary to ensure the code’s functionality aligns with its intended purpose, he adds.

Also Read: Mint Explainer: How banking staff is paying price for digitalization-led growth

Chatbots are another AI-driven innovation that transforms customer and internal interactions, says Leukert. They can handle routine requests such as reporting a lost credit card or increasing a credit limit with minimal human intervention. This not only improves service efficiency but also enhances the customer experience, he explains. 

Finally, AI's role in content analysis and recommendation is invaluable, he says. By processing vast amounts of data—from financial trends to geopolitical developments—AI helps banks offer timely and relevant advice to clients. This capability accelerates decision-making and enhances the quality of recommendations provided to customers, he adds.

Leukert acknowledges that despite these advancements, challenges persist. Ensuring data privacy and security, managing the complexity of AI integration and maintaining transparency in AI-driven decisions are critical concerns, he affirms. Addressing these issues is essential to maximise the benefits of AI while safeguarding user trust and operational integrity. 

“We adhere to a crucial principle: We always have humans in the loop. We ensure that all content generated by large language models (LLMs) is reviewed and approved by a human before being released," says Leukert.

Additionally, he says, Deutsche Bank implements stringent governance and control measures, particularly in banking, which may slow down deployment but ensures safety and compliance. Even prototypes must pass these controls before being used productively. AI tools like chatbots are tested rigorously internally before external release.

Deutsche Bank, according to Leukert, began with a strategy based on "future technology forecasts" 4-5 years ago.

"It's now paying off," he says.

Using the cloud

Leveraging the cloud, the bank uses infrastructure- and platform-as-a-service "where we monitor the vendors who supply the service for us, as opposed to outsourcing, where you own the service and have vendors delivering according to your SLAs (service level agreements)," says Leukert. 

The idea is that “Deutsche Bank should focus on banking and not on running data centres and keeping data centres up to date with the latest technology infrastructure." This ensures that all applications that are relevant internally but more important that customers are leveraging the latest hardware and software, he adds.

Deutsche Bank partners with Google for its banking applications and Microsoft for end-user services to ensure resilience and financial stability. It uses Google Cloud to run Autobahn FX, its primary trading platform for foreign exchange currencies and precious metals like gold and silver.

This separation mitigates risks such as technological issues or natural catastrophes, allowing banking operations to continue even if end-user services face disruptions, according to Leukert. A distributed cloud also helps the bank comply with global regulations like data localisation.

According to Leukert, the partnership strategy helps Deutsche Bank to improve its “industry resilience," increase its response time when addressing customer problems, and have the “elasticity to account for geopolitical events, or a massive increase of interest rates to fight inflation, or people working from home that is the new normal today -- this is where the cloud helps you scale your network capacity when needed."

Also Read: Soaring profits are no reason to go slow on banking reform

That said, the return on investment (ROI) for generative AI remains a contentious topic, with ongoing debates among experts and consultancy firms. To be sure, Deutsche Bank is also collaborating with Publicis Sapient to accelerate the adoption of GenAI. The bank made these investments to improve its return on equity and reduce its cost-to-income ratio (CIR) to under 62.5 percent by the end of 2025, according to an 18 March report in HFS Research.

"This partnership will see Deutsche Bank develop its AI agenda with the goal of driving new business models to boost revenues—and it’s a model other traditional banks should follow," the report says.

Assessing the impact

Nevertheless, the challenge lies in measuring ROI for technologies that are still evolving.

"Currently, we are in the hype cycle phase, where the practical benefits of AI are still being assessed. We are cautious about fully deploying AI solutions if they cannot show a positive impact on their profit and loss (P&L) statements. For instance, Microsoft’s Copilot may enhance productivity in creating PowerPoint presentations, but if the cost per use outweighs the savings from manual creation, it may not justify widespread adoption," Leukert says. 

Also Read: GenAI is a force multiplier, and India will be the testbed: McKinsey CTO

He adds that AI can streamline processes such as in content generation but may not always translate directly into P&L gains. In other cases, AI-driven efficiency can lead to workforce reductions or shifts, focusing on retraining employees rather than layoffs.

The goal, according to Leukert, “is to balance technology costs with operational improvements, adjusting the workforce as needed to maximise overall productivity and financial performance."

Global talent

Towards this end, the bank is “strengthening our engineering skills, and engineering culture internally." Hence, while the total workforce “has kind of more or less kept stable, the ratio of engineers has massively gone up," he says.

"We are headquartered in Germany but have access to global talent pools. We have massively kind of grown our presence in India," he points out.

The India tech centre was established in 2015, and the bank has since “grown from about 100 people to close to 8,000, or about 40% of our employees who work in our technology, data and innovation division."

The next step, according to Leukert, is “to continue allocating global responsibilities into India and providing a service across our units -- from the US and Europe into Asia."

Leukert says that Deutsche Bank has already started its so-called "Leadership Accountability" in India. 

"As we speak, our global head of procurement IT, our global head of legal IT, and our global head of Treasury IT are sitting in India, providing services globally, which was different from the past where most of the leadership was in Frankfurt, London, and a little bit in New York," says Leukert.