As banks and financial institutions rapidly embrace artificial intelligence, software testing and quality assurance are taking on unprecedented importance.
In a future where AI underpins every customer interaction and back-end operation, QA is no longer a supporting function, it’s the backbone of trust, security, and innovation, according to McKinsey & Company.
In one of the firm’s recent reports, called ‘Building the AI Bank of the Future’, senior partner Renny Thomas outlines a vision of a transformed financial ecosystem.
One of the report’s foundational insights is that most banks today still operate in ways “that date back to the last century,” with legacy architectures and disconnected data stacks hindering progress. As AI redefines what’s possible, the ability to reliably test, deploy, and monitor systems at scale becomes mission-critical.
“AI-first banks will need to develop modular, flexible, cloud-based core technology platforms,” stated Mumbai, India-based Thomas, who leads McKinsey’s banking work across Asia–Pacific and works with a host of financial services clients amid the rapid growth of Asia’s banking sector.
These platforms must support high volumes of automation, real-time analytics, and rapid experimentation, but only if robust QA processes are in place, he stressed.
In a world where algorithms make credit decisions and monitor for fraud, faulty software can lead to reputational damage, regulatory penalties, and direct financial losses.
Banks are beginning to recognise that AI capabilities will require far more rigorous validation protocols than traditional IT.
“Banks will need to develop rigorous model-risk-management and monitoring capabilities, especially as regulators and watchdog organizations push for more model accountability and transparency,” Thomas stated.
Testing these models in varied, complex environments, and ensuring reproducibility and auditability, is no small task.
The future of banking, as envisioned by the McKinsey report, rests on two complementary foundations: digitised operations and AI-driven decisioning. That dual mandate intensifies the already enormous pressure on QA teams.
“Only a bank that is digitised to the core can fully benefit from embedding AI across all of its operations,” Thomas asserted.
That digitisation introduces a vast array of interdependencies, such as APIs, cloud-native microservices, machine learning models, third-party tools, all of which must be tested not just in isolation, but as a dynamic whole.
According to the report, “the transition to an AI-first bank will require significant investment over several years.” That investment isn’t limited to hiring data scientists or upgrading infrastructure; it must also include test automation, observability tools, synthetic data generation, and intelligent QA orchestration. Without these, banks risk building brittle systems that fail under pressure.
Thomas noted that future-ready banks will not only have to develop but also integrate new processes to safeguard quality at speed.
“AI-first banks will require a robust data architecture and data management capability,” he said. That means QA teams must now extend their mandate beyond code and UI testing, incorporating data validation, AI behavior modeling, and continuous performance monitoring into their toolkits.
Trust is the currency of financial services, and in an AI-first world, trust must be engineered into every release.
Thomas underscored this point: “They must also make foundational investments in next-generation capabilities, including machine learning, MLOps, and a modern data architecture, as well as in talent and ways of working.”
Modern QA must be able to test not only whether software works—but whether the outcomes it produces are fair, explainable, and compliant.
“AI-first banks will need to build, test and deploy hundreds of AI applications across the organisation.”
– Renny Thomas
While the vision of an AI-first bank is ambitious, the risks of underinvesting in QA are even greater. “Most banks are still in the early stages of their transformation journeys,” Thomas acknowledged. And without embedding QA deeply into that journey, transformation may falter before it scales.
There’s also the reality that traditional banks must move fast to keep up with digital-native challengers who have quality baked into their DevOps cultures.
“AI-first banks will need to be able to deploy code into production hundreds or even thousands of times per month, depending on the business need,” Thomas said. Such velocity is only sustainable with automated, scalable, and AI-augmented QA systems.
Thomas and the McKinsey team recommend a reimagining of the operating model itself. “AI-first banks will need to build, test and deploy hundreds of AI applications across the organisation,” he said.
Each of those applications, from customer service bots to real-time risk scoring engines, must go through rigorous scenario-based testing, adversarial testing, and long-term drift monitoring to avoid system degradation.
The underlying message from McKinsey is clear: excellence in QA is not a bottleneck to innovation, it’s the enabler. “Banks that go all in on AI and recognise the structural capabilities and enablers required to embed it across the organisation can generate a significant and sustainable competitive advantage,” Thomas said.
In practical terms, that means QA must be present from the moment new products are conceived. Banks should embrace shift-left testing strategies, embedding test engineers and model validators early in the development pipeline.
Observability tools
Observability tools must be built into every component, enabling teams to monitor not just failures, but anomalies, delays, and unintended outputs.
Thomas emphasised the need for “integrated and modular technology stack[s], including data platform and cloud infrastructure, machine learning tools, and MLOps frameworks.”
Each layer introduces its own testing challenges, whether that is simulating real-time traffic spikes on APIs or validating the fairness of AI-generated credit scoring.
The future of banking QA, therefore, lies at the intersection of disciplines: software engineering, data science, cybersecurity, compliance, and customer experience. The days of QA as a post-development checkpoint are over. Thomas made that clear: “AI-first banks will operate as technology companies with banking licenses.”
Ultimately, McKinsey’s report paints a picture not just of transformation, but of an entirely new discipline emerging within banking: intelligent assurance. It’s a shift from “does it work?” to “does it perform, adapt, and remain trustworthy over time?”
As AI becomes embedded in everything from account opening to fraud monitoring, QA professionals will be tasked not just with preventing bugs, but with safeguarding fairness, compliance, and customer confidence. And that, according to Thomas, is not a nice-to-have, it’s the foundation of the future bank.
Singapore leading the pack
Despite many banking giants in Wall Street and the City of London rapidly embracing GenAI, the financial hub of Singapore is generally seen as being ahead of the rest when it comes to investment and adoption of AI.
In fact, earlier this year Singapore’s DBS Bank confirmed plans to lay off around 4,000 people in the next three years as artificial intelligence tools and solutions will take over the work that is currently carried out by humans.
The regional banking giant is currently testing and trialling a host of AI-powered software systems and platforms.
While thousands of jobs are expected to go, the bank did point out that the further integration of AI technology does mean around 1,000 new, related jobs are expected to be created.
Last year, DBS disclosed it had, at the end of 2024, deployed around 800 AI models across 350 use cases, and expected the measured economic impact of these to exceed S$1 billion before the end of 2025.
The announcement comes as current deputy chief executive Tan Su Shan is preparing to take the top job next month.
The tech push is remarkable as it makes DBS one of the first major banks globally that has set out specific plans on how and when AI may affect its operations in such detail.
Many other banks, most notably in the U.S., may have indicated AI is coming in fast, with junior roles likely to be replaced by the relatively young technology, but detailed plans and specific strategies have hardly been presented, to date.
DBS’ announcement fits into a wider trend within Singapore’s financial services space to embrace AI technology, largely encouraged by the Monetary Authority of Singapore (MAS), the micro-state’s financial services watchdog.
At the end of last year, MAS announced it would commit S$100 million, or close to $75m, to support the island-state’s banks and other financial services firms to design, test and build capabilities in artificial intelligence technologies.
The capital was meant to speed up the advancement of AI-related innovation and adoption in financial services, MAS said.
With the advent of Generative AI, financial institutions have embarked on initiatives to map the technology’s opportunities and risks, and have begun piloting it across a range of use cases, it stressed.
“Nevertheless, the level of AI-readiness and adoption varies hugely across financial institutions in Singapore,” the regulator warned.
“MAS will therefore bolster financial institutions’ development and deployment of AI technologies in Singapore,” it said.
The watchdog firmly believes that Singapore has “the potential to become a centre of excellence for anchoring AI capabilities, such as in the development of applications, as well as testing and deployment of AI solutions for the financial sector.”
MAS will support financial institutions in establishing AI innovation centres in Singapore for a range of functions including AI model building and training, deployment of AI models for high-impact use cases, governance and risk management, as well as testing and monitoring.
It also plans to develop AI platforms to address industry-wide use cases.
“There are strong prospects for the financial industry to apply AI to solve industry-wide problems beyond what each financial institution can do individually,” MAS clarified.
This involves the development of frameworks and platforms for policies and protocols that enable secure and privacy-protected data exchange where financial institutions can collaborate on industry-wide use cases, the regulator concluded.
NEW EVENT
Why not become a QA Financial subscriber?
It’s entirely FREE
* Receive our weekly newsletter every Wednesday * Get priority invitations to our Forum events *
REGULATION & COMPLIANCE
Looking for more news on regulations and compliance requirements driving developments in software quality engineering at financial firms? Visit our dedicated Regulation & Compliance page here.
READ MORE
- Banks confront rising agentic AI testing challenge
- Testing turns ‘reactive’ as documentation lags
- Inside UBS’s landmark testing challenge
- Bank of England raises the testing bar for frontier AI
- Tricentis, Tabnine and Snyk: the latest vendor and product news
WATCH NOW
