Lead with impact: Turning AI into business decisions in financial services

The use of advanced analytics and machine learning in financial services has never been the problem. The industry was an early mover and built real capability. What's changed is the capability itself, and with it, the gap between what AI can now do and what most organizations are actually capturing from it.

According to a Broadridge survey, AI adoption in financial services is up 49% year over year. But value creation is up just 13%. The gap between those two numbers isn't a technology problem - it’s a people problem. The models' power doesn't reach the right people in the organization, at the scale of the business.

At the same time, the regulatory environment is accelerating. SR 26-2 is here. The NAIC pilots are expanding across states. DORA and the EU AI Act are live for European enterprises. Getting ahead of this moment requires treating governance not as a constraint on innovation but as the infrastructure that makes it sustainable.

Four imperatives define what that looks like in practice right now:

The governance imperative: Accelerate innovation, don’t slow it down

The governance approaches designed for SR 11-7 were built for a different era -deterministic models, slow-moving risk, relatively contained scope. That's not the AI financial services organizations are deploying today. GenAI is non-deterministic, agentic systems make autonomous decisions in real time, and models drift in minutes, not months. What worked then requires fundamental rethinking now.

What modern governance looks like in practice:

• Governance built in, not bolted on. Leading organizations have moved from governance as a checkpoint to governance as infrastructure, so it’s automated, continuous, and embedded in the workflow from the start. Manual evidence collection and audit backlogs are solvable problems, not inevitable ones.
• Navigating an accelerating regulatory agenda. SR 26-2 updates MRM guidance for the first time in fifteen years, but GenAI and agentic AI are explicitly out of scope. Organizations are being asked to govern a rapidly evolving landscape that the frameworks don't fully cover. What's needed is the flexibility to adapt as guidance catches up to the technology.
• Governance as a value driver. The cost of poor governance isn't just regulatory exposure. It's the audit backlogs, redundant work, and compounding technical debt that slow everything else down.

The application imperative: Closing the AI last mile

Financial services organizations have built models that could transform credit risk, fraud detection, underwriting, AML/KYC, and market surveillance. Most of that potential is still sitting unrealized. Not because the models aren't good enough, but because they don't reach the people making decisions. The last mile between a model in production and a business outcome is an application problem, not a modeling problem.

What closing that gap looks like in practice:

• Delivering insights at scale. Providing access to models through dedicated, secure, governed, and scalable applications puts insights in the hands of the analysts, risk teams, and relationship managers who need them.
• Governance doesn't stop at the model. The criticality of financial decisions and the sensitivity of the data involved mean AI-powered applications require the same rigor as the models underneath them. Scalability and user experience are part of the compliance picture, not separate from it.
• Putting humans in the loop. The highest-stakes decisions in financial services - fraud detection, credit adjudication, risk assessment - require human judgment informed by robust data. Workflows that bring humans and AI together with the right guardrails are where the real transformation happens.

The innovation imperative: Agentic is changing how financial services build and operate

According to KPMG, 99% of organizations plan to deploy AI agents. Only 11% have done so. The gap reflects the genuine complexity of innovating in a high-stakes, heavily regulated environment, and it's also where the largest opportunity sits. Agents that monitor transactions, streamline underwriting, and support relationship managers aren't hypothetical. Organizations are deploying them in production right now.

What that requires to get right:

• A new way to build. Coding agents are changing the economics of software development. The opportunity isn't just faster pilots. It's building for scale and long-term maintainability without accumulating a new layer of ungoverned technical debt.
• Infrastructure that connects. Agents don't deliver value in isolation. They have to connect to enterprise data and systems to matter. Financial services organizations are already paying the price of years of AI fragmentation. A unified approach, open to innovation and enabling consolidation, is what makes agents credible at enterprise scale.
• Real-time governance for autonomous systems. Agents make decisions and collaborate with humans in ways that require a fundamentally different supervision model. Real-time governance, dynamic guardrails, and clear human oversight aren't optional; they're what make agentic systems credible in regulated contexts.

The ethical imperative: A framework built for what’s actually being deployed

Responsible AI frameworks were designed for narrow, predictive models. They strain under GenAI and break under agentic systems because the technology has outpaced the frameworks. As financial services organizations deploy agents that make autonomous decisions in high-stakes contexts, the question isn't just whether they're compliant. It's whether the ethical and operational frameworks governing them are built for what they're actually doing.

Reid Blackman, founder and CEO of Virtue and one of the leading voices on AI ethics in regulated industries, argues that financial services organizations need to fundamentally rethink their responsible AI frameworks for the agentic era, not just update them. He makes that case in his new book, just out this month.

The conversation happening in New York on May 19

Rev New York brings together leaders from across financial services who are navigating these imperatives and sharing what they've learned, including what they'd do differently.

Reid Blackman opens the day with that ethical foundation. David Palmer, author of SR 11-7, closes with where governance is heading after SR 26-2. That should tell you everything about the level of conversation in the room.

Whether you're scaling model risk management, deploying agentic systems, or trying to close the gap between AI investment and business impact, the conversations at Rev are designed to shape your strategy, sharpen your execution, and give you an honest picture of what actually works. Register at https://rev.domino.ai/new-york