Laurent Giraid is a seasoned technologist specializing in Artificial Intelligence and its ethical implementation within the financial sector. With years of experience navigating the intersection of machine learning and banking infrastructure, he offers a deep perspective on the latest industry trends where AI has moved from a speculative luxury to a functional necessity. In this discussion, we explore the findings of the latest industry research involving over 1,500 senior executives, delving into the shift from simple pilots to enterprise-wide scaling. We examine the critical infrastructure upgrades required to sustain this momentum, the growing importance of agentic AI, and the regional nuances that dictate how financial institutions are evolving globally.
With AI adoption now nearly universal across the industry, how can institutions move beyond basic deployment to create a unique competitive advantage? What specific performance metrics or operational outcomes should leaders prioritize to prove their AI implementation is actually providing a differentiator?
We have reached a fascinating tipping point where only 2% of financial institutions globally report no use of AI, meaning the mere presence of the technology is no longer a badge of honor. To truly stand out, leaders must shift their focus toward operational complexity and the seamless integration of AI across the entire value chain rather than keeping it in isolated pockets. I suggest prioritizing the “innovation lever” metric, as 43% of your peers already identify AI as their primary driver for evolution. Success is no longer measured by the number of pilots but by how effectively a firm can scale these tools responsibly to improve the 60% of capabilities that were reportedly enhanced over the last year. The real differentiator lies in moving from “experimental” to “reliable,” ensuring that every automated outcome is backed by a framework that both the board and the regulators can trust.
Risk management and fraud detection are currently the most common areas for AI integration. How are these tools evolving from simple detection to predictive prevention, and what steps are necessary to ensure these systems do not generate excessive false positives that disrupt the customer experience?
Currently, 71% of institutions are focusing their AI programs on risk management and fraud detection, marking it as the most mature sector of deployment. The evolution we are witnessing involves moving away from reactive rules-based systems toward predictive models that analyze vast amounts of data in real-time to stop threats before they materialize. However, the sensory experience for the customer can quickly turn from “secure” to “frustrating” if false positives are not managed with surgical precision. To balance this, firms are increasingly looking at document intelligence and data analysis—which also see 71% and 69% engagement respectively—to provide a more holistic view of the customer. By refining these models, we can ensure that the 69% of firms using AI for customer service assistants can resolve issues with a human-like touch, rather than a robotic barrier.
Roughly nine in ten firms are prioritizing infrastructure modernization to support AI scaling. Which specific technical debt or legacy core banking issues are currently the biggest bottlenecks, and how should a CIO balance immediate AI deployment with the long-term need for cloud-native data platforms?
The reality is that AI is only as capable as the systems underneath it, which is why 87% of institutions are planning significant modernization investments over the next 12 months. Many CIOs feel the crushing weight of legacy core banking systems that act as a bottleneck, preventing the fluid data movement required for high-speed machine learning. There is a palpable pressure to deliver immediate AI results, yet the foundational layer of cloud adoption and data platform modernization must be the priority to avoid a total system collapse under the weight of new demands. It is a delicate act of building the airplane while flying it, but the data shows that 54% of institutions are now using fintech partnerships as a shortcut to bypass these legacy hurdles. This collaborative approach allows for rapid deployment while the internal team works on the long-term transition to a more agile, cloud-native architecture.
Many regions, particularly the United States and Singapore, are facing significant talent shortages in the specialized AI space. Beyond aggressive recruiting, what internal training strategies or fintech partnerships have you seen succeed in bridging this skills gap while keeping project costs under control?
The talent war is incredibly intense right now, with 43% of institutions citing a lack of expertise as their primary obstacle, and that figure climbs to a staggering 54% in Singapore. In markets like the US and Japan, where 50% of firms are struggling to find the right people, the atmosphere in HR departments is one of focused urgency. To combat this without blowing the budget on astronomical salaries, successful firms are leaning heavily into the “partnership-as-a-service” model. By turning to fintech partners—the preferred strategy for 54% of respondents—banks can “rent” the expertise they cannot yet “buy” or “build.” This provides an immediate infusion of high-level skill while internal teams are upskilled through the shared management of these sophisticated projects.
As agentic AI systems begin to handle multi-step tasks and autonomous decision-making, how must internal governance frameworks change? What practical methods can be used to ensure these models remain explainable and auditable, especially when regulators demand transparency for every automated financial outcome?
With 63% of institutions already piloting agentic AI, we are entering a realm where software makes consequential, autonomous decisions that go far beyond simple data sorting. This shift requires a radical overhaul of governance because the “black box” approach is no longer acceptable to regulators or the customers who demand reliability. We must implement “explainability by design,” ensuring that every multi-step task performed by an agent can be traced back to a logical, auditable trigger. It is about creating a paper trail for the digital age, where the logic of the algorithm is as transparent as a traditional ledger. If we cannot explain why an AI denied a loan or flagged a transaction, we risk not just a regulatory fine, but a total loss of the reputational trust that takes decades to build.
Regional priorities vary significantly, with some markets focusing on rapid financial inclusion while others emphasize cautious incremental change. How can global institutions adapt their AI strategies to respect these local market constraints while still maintaining a consistent, secure enterprise-wide technology standard?
The global landscape is a study in contrasts, from Vietnam’s aggressive 74% active AI deployment rate to Japan’s more measured 39%. In Vietnam, the drive is fueled by a visceral need for financial inclusion and rapid payment processing, whereas Japan’s caution stems from a deep-rooted cultural preference for incremental change and the constraints of legacy infrastructure. To navigate this, a global institution must adopt a “glocal” strategy: maintaining a rigid, high-standard enterprise core for security and data integrity while allowing local teams the flexibility to deploy specific use cases. For instance, while Singapore might prioritize a 50% increase in personalization spending, a branch in a more conservative market might focus solely on the back-office efficiency of document management. This allows the firm to respect local pace without compromising the overarching goal of becoming a digital-first entity.
What is your forecast for AI adoption in financial services?
My forecast is that we are moving toward a period of “invisible AI,” where the technology becomes so deeply embedded in the financial value chain that we stop talking about it as a separate initiative. Within the next few years, the 87% of firms currently modernizing their stacks will have completed their transition to cloud-native environments, making AI-driven personalization and autonomous workflows the standard operating procedure. We will see a shift where the competitive gap widens significantly between those who mastered governance early and those who treated it as an afterthought. Ultimately, the industry will be defined by institutions that provide financial services that are not just faster, but more secure and deeply personal, proving that they have successfully crossed the tipping point from experimentation to essential reliability.
