The Collateral Advantage: AI, Data and the Operating Model Ahead

Collateral administration is entering a transformative phase. As markets grow more complex and liquidity becomes increasingly valuable, the function is evolving beyond its traditional roots. What was once a back-office process focused primarily on control is now a strategic driver for resilience, efficiency and optimization. For asset managers and institutional firms carrying collateral-intensive portfolios, the limits of legacy operating models are no longer a future concern, they are a present operational constraint.

As eligibility rules tighten, settlement speeds accelerate and margin requirements grow, the collateral administration function is being forced to operate in real-time, connected environments that fragmented infrastructure was never built to support. Innovation and artificial intelligence (AI) are central to this shift, offering the tools to simplify complexity and sharpen decision-making. Sean Lynn, Director of BNY’s Collateral Administration business, explains how these technologies are shaping the future.

What market dynamics are driving the next phase of change in collateral administration?

A number of forces are converging at once, and that is what is driving the next phase of change in collateral administration.

Volumes continue to rise, eligibility rules are becoming more complex, and clients expect greater speed and transparency. At the same time, firms remain focused on capital and liquidity efficiency. Taken together, those pressures are exposing the limits of more traditional operating models.

Where firms still rely on fragmented data, manual processes and legacy technology, it becomes much harder to keep pace. Those models are often slower, less connected and more difficult to scale.

This is why I see the current moment as more than an opportunity for incremental improvement. It is a broader shift in how collateral administration needs to operate in an increasingly real-time, connected and data-driven environment.

What is the role of innovation and AI in the future of collateral administration?

The most significant opportunity is that AI can help move collateral administration from a largely reactive function to one that is more predictive and forward-looking.

Historically, much of the focus has been on processing activity, managing exceptions and ensuring timely execution. That foundation remains important, but AI creates the potential to identify issues earlier, improve prioritization and support faster, better-informed decisions.

Collateral administration is well suited to this kind of innovation because the workflows are high-volume, time-sensitive and often exception-driven. There is a clear opportunity to use AI to support triage, routing and decision-making across large and complex datasets.

That said, technology only creates value when it is embedded in the operating model itself. It cannot simply sit on top of existing complexity. Firms also need stronger connectivity across platforms and data that is more accessible and usable. That is what turns innovation into something practical and scalable.

What are the most meaningful operational benefits of applying AI?

Efficiency is the most immediate benefit, but the broader value lies in reducing friction across the end-to-end process.

By automating routine activities such as validation, reconciliation and certain forms of exception handling, firms can improve speed, consistency and operational resilience. As a result, they can then free up experienced teams to focus on areas where judgement, oversight and client engagement add the most value.

That changes the operating model in a fundamental way. It is not only about doing the same work faster; it is about improving how the work is performed and where people can contribute most effectively.

Another important benefit is visibility. If AI can surface anomalies earlier and help prioritize exceptions more effectively, it can strengthen control alongside efficiency.

At the same time, interoperability remains a real challenge. If AI cannot operate effectively across internal systems, vendor platforms and client tools, it becomes much harder to capture its full potential.

How can AI support collateral and liquidity decision-making?

This is one of the areas where AI has the greatest long-term potential.

Collateral decisions are inherently complex. Firms are balancing asset availability, eligibility requirements, haircuts, concentration limits, funding costs and liquidity needs, all while market conditions continue to shift. Managing those variables manually, and at speed, is increasingly difficult.

AI can help evaluate those inputs in real time and support more effective optimization decisions. That can lead to smarter asset allocation, reduced inefficiencies and a stronger overall approach to liquidity management.

Using AI in this way can also help enable firms to be more forward-looking. If collateral demand and settlement pressures can be identified earlier, firms are in a stronger position to respond effectively. In fast-moving markets, that visibility can create a meaningful advantage.

What does successful AI adoption require?

Successful AI adoption starts with the right foundation.

Data quality is critical. If data is inconsistent, incomplete or siloed, even the most advanced tools will struggle to deliver reliable outcomes. For that reason, the underlying data environment has to be addressed early.

It is also important to be deliberate about where AI is first applied. The most effective approach is usually to begin with use cases where operational friction is clear and measurable, allowing firms to demonstrate value early and build momentum over time.

Human oversight also remains essential. In collateral administration, governance, transparency and accountability are fundamental. The strongest operating models will be those in which technology enhances human decision-making and execution, rather than attempting to replace them.

How will innovation and AI reshape collateral administration over the next few years?

Collateral administration is moving steadily toward a more real-time and data-driven operating model.

AI will help firms anticipate needs earlier, optimize liquidity more effectively and respond faster as conditions change. That will improve efficiency, but it will also reshape how firms think about oversight, control and client service.

More broadly, I expect the industry to continue moving away from fragmented models and toward platforms that are more connected, more intelligent and better able to support real-time decision-making.

Vendors will have an important role to play in that transition as well. Progress will depend not only on better interfaces, but also on stronger connectivity, more effective APIs and architecture capable of supporting dynamic, AI-enabled workflows.

Ultimately, the firms that lead in this space will be those that view AI as part of a broader evolution in the operating model. It is not simply another tool; it is part of building a function that is more adaptive, more efficient and better prepared for the market ahead.