Financial institutions (FIs) are a cornerstone of economic and societal stability, but their size and complexity make it difficult for them to change with markets and technologies. FIs host vast quantities of sensitive data, however very little of that data is put to effective use despite a huge regulatory burden and the potential value and insight that data could afford.

This spring, we’re sharing a series of blog posts exploring AI technology hype and its applies to the risk industry. Missed our earlier posts? Read The AI Hype Machine, Part I and Part II, then come back. Today we explore the current state of financial risk technology.

Legacy Needs Technology

Managing risk is about knowing, and the pressure on financial institutions to know has never been greater.

On the quest to complete knowledge, FIs have become prodigious hoarders of information. “Data lakes” is a fitting term for the comically enormous quantities of information these firms house on their servers. They’re vast, churning, and heterogeneous bodies, their borders advancing as terabits fall daily from the heavens. Customers onboard, traders trade, accounts fluctuate more, and every moment, and it won’t ever end. And regardless of the difficulty, organizations must know.

They must know because it’s their civic obligation—and because failure to know can have devastating consequences. Fraud can topple economies, and regulatory penalties can mean billions in lost revenue. To be a FI post-recession is be in a perpetual existential crisis over a sisyphean epistemological paradox: your survival depends on attempting omniscience.

With stakes like these, one would think FIs bought or built risk management systems integrated with every applicable advancement. With so much data on hand, one would assume that their data lakes were charted by the best AI applications.

It would be reasonable to think that, but you would be wrong.

Updating Atlas

Financial institutions move slowly—and with good reason. They are foundational to the continued health and growth of modern society, and, as the aphorism goes, it’s hard to build on shifting sand.

But beyond reasons of consistency, the glacial pace of change is also a consequence of heavy restraints. FIs can be huge: trillion-dollar, millions of customers huge. Their operations are supported by employees that can number in the hundreds of thousands, enmeshed in a myriad of interconnected workflows. Serious interruptions to this symphony of money, math, and people could spell a catastrophe. Just taking their sheer size and complexity into account, it’s no wonder that systemic overhauls map to generations.

The burden of size is intensified when regulations are brought into the equation. FIs not only have to deal with the logistical headache of infrastructural overhaul, but also the task of tuning new systems to the stringent and costly requirements of regulators. The fact that some of the world’s most powerful institutions have ancient mainframes under the hood may be hard to swallow, but, for them, it’s a just better bitter pill.¹

Regardless of an ill-stacked deck, FIs must upgrade. They need to have a single view of their customers. They need to be able to resolve identities and spot fraud. They need to be able to leverage the insight floating in their data lakes. Regulations demand it, and the penalties can have nine zeros. Silicon Valley is beginning to lick its chops, and far off in the distance they can see what financial innovations like Bitcoin could mean for the future.

If FIs are going to get modern about risk, they’re going to need AI to do it.

Learn more

With so much hype and disappointment surrounding AI, what can it do for risk technology today? Check back next week to find out, or download the complete Honest Guide to AI for Risk now.

---

^{1 Prattle, “Why Financial Institutions Struggle to Modernize…and Why They Have to”}