Today, we’re joined by Priya Jaiswal, a recognized authority in Banking, Business, and Finance, whose expertise in market analysis and international business trends provides a unique lens on the financial sector’s technological evolution. Our conversation will explore the tangible strategies major banks are deploying to integrate artificial intelligence, moving beyond the hype to remake core operations. We’ll touch on how these institutions are navigating the significant investments required, the key hurdles to achieving a true return on investment, and the internal metrics that signal growing confidence among industry leaders. Finally, we’ll examine what it takes to build a foundational AI platform that empowers every employee, transforming how work gets done across the entire enterprise.
BNY Mellon is integrating advanced models from Google and OpenAI into its proprietary platform. Beyond technical integration, what are the key strategic steps for deploying this technology at scale and successfully remaking core processes in new and exciting ways? Please provide a specific example.
The real challenge isn’t just about plugging in a powerful model like Gemini; it’s about fundamentally re-architecting the work itself. When a bank like BNY Mellon talks about remaking processes, they’re moving beyond simple automation. The first strategic step is to identify a core process ripe for reinvention, not just improvement. Then, you must create a cross-functional team of technologists, business-line experts, and compliance officers to map out not just the technology but the human workflow around it. For example, consider post-trade settlement. Traditionally, it’s a series of checks and balances. With AI integrated into a platform like Eliza, you could build a predictive system that analyzes incoming trade data in real-time, flags potential settlement failures before they even happen, and suggests corrective actions. This shifts the entire process from a reactive, manual reconciliation to a proactive, automated risk management function, which is a truly new and exciting way to operate.
Citi is assessing AI’s role in complex processes like “know your customer” and loan underwriting. Can you walk us through how AI re-engineers one of these workflows step-by-step and what specific new efficiencies are being unlocked that were previously unimaginable?
Let’s break down the “know your customer” or KYC process. Historically, this has been an incredibly manual and time-consuming workflow. An analyst would collect documents, manually verify identities, sift through public records, and make a judgment call. The process was slow and could be inconsistent. Now, with AI, the first step is automated data ingestion, where the system scans and extracts information from various documents instantly. Next, an AI agent can analyze vast, unstructured datasets—news articles, social media, corporate filings—to build a comprehensive risk profile in seconds, something a human could never do at that scale. Finally, instead of a one-time check, the AI provides continuous monitoring, alerting the bank to any changes in a client’s risk profile. The efficiency that was unimaginable just a few years ago is this shift from a static, snapshot-in-time approval to a dynamic, living risk assessment that operates 24/7.
With firms like Bank of America investing hundreds of millions in AI, there is still a challenge in reaping the technology’s full benefits. What are the primary hurdles to achieving ROI, and what practical steps should executives take to apply these tools more effectively across the company?
The single biggest hurdle is moving from isolated pilot projects to true enterprise-wide adoption. A bank can pour hundreds of millions into 20 different AI initiatives, as Bank of America has, but if those projects remain siloed within specific departments, you’ll never achieve transformative ROI. The technology becomes a collection of shiny objects rather than a foundational capability. The primary practical step executives must take is to focus on enablement. This means investing in a common, underlying platform and set of tools that can be used across the company. The goal is to make AI accessible not just to data scientists but to business analysts, legal teams, and HR professionals. When the entire organization sees AI as a tool to augment their daily work, that’s when you start seeing the widespread efficiency gains that justify the massive initial investment.
Executives at Morgan Stanley and Goldman Sachs express growing confidence in AI’s potential. What specific milestones or internal metrics are they likely tracking to justify this optimism, and how does reshaping a few key operating processes ultimately translate into broader enterprise growth?
Their confidence isn’t based on vague potential; it’s rooted in concrete, process-level metrics. For the six operating processes Goldman Sachs is targeting, they are almost certainly tracking things like a reduction in manual processing hours, faster client onboarding times, or improved accuracy in risk model outputs. These are tangible, quantifiable wins. The magic happens when these individual process improvements start to compound. For instance, if you can make your software engineering more efficient with AI, you can deploy new client-facing products faster. If you streamline compliance checks, you can onboard more clients in the same amount of time. Reshaping these core processes frees up capital and, more importantly, human talent. This “freed up capacity,” as David Solomon put it, can then be reinvested into growth areas, turning internal efficiency directly into a competitive advantage and a driver for enterprise growth.
For AI to significantly trim costs, broad employee enablement is essential. What does a successful foundational AI platform look like from a CIO’s perspective, and how can they ensure employees in non-technical departments like legal and HR actively leverage it to change how they work?
From a CIO’s perspective, a successful foundational platform isn’t just a collection of APIs; it’s a secure, user-friendly ecosystem. It must have robust governance and security controls baked in, especially in a regulated industry like banking. Critically, it should offer low-code or no-code tools, allowing non-technical users to build their own simple AI agents or workflows. To ensure adoption, the CIO needs to spearhead a cultural shift. This involves creating “AI champions” within departments like legal and HR who can demonstrate practical use cases. For example, the legal team could use the platform to quickly analyze contracts for specific clauses, while HR could use it to draft initial job descriptions or parse thousands of resumes. By providing the tools and showcasing clear, immediate benefits for their specific jobs, you empower them to change how they work, which is the only way to realize those potential 20% cost savings across the board.
What is your forecast for AI adoption in banking?
My forecast is that over the next two to three years, we will see a clear separation between banks that treat AI as a series of disconnected technology projects and those that successfully embed it as a core enterprise capability. The conversation will shift from “if” to “how effectively.” The leaders will be those who have not only invested in the technology but have also invested in the cultural and operational changes required to support it. Banks that succeed in building a foundational platform for broad employee enablement will unlock compounding efficiencies, while laggards will find themselves with expensive, underutilized tools and a widening competitive gap. The true measure of success won’t be the number of models in production, but the degree to which AI has fundamentally changed the way every single employee works and thinks.
