In 2023, no other technological innovation has seized the global spotlight in the way that generative AI has. In the same way that the cloud opened up new possibilities for both consumers and businesses, generative AI has vast potential to transform all major verticals, including financial services. According to McKinsey research, generative AI could add the equivalent of $2.6 trillion to $4.4 trillion to the global economy annually. And from our point of view, this potential extends to custody and clearing.
But the question is – how? And how soon? And to what end?
Despite all of the frenzy around the future applications of generative AI, there are a number of very tangible ways in which generative AI is already in the process of being implemented today.
Real-time client service
Custody and clearing firms house vast repositories of data on behalf of their clients – from individual investor account documents to historical trading activity. This data is already the critical ingredient for any fintech or advisor to build powerful large language models (LLM).
The challenge with public AI (think ChatGPT) is that these models aren’t tailored to a regulated fintech or advisor, or even to the unique guardrails or terminology that exist within the financial industry. A “bond” in the context of investing is dramatically different from a social bond between two friends, for example.
But – what if a fintech or advisor could custom-build their own LLM using data from their custody and clearing provider?
Practically speaking, generative AI will be able to replace the basic, high-volume interactions between a customer and their customer support team – freeing up customer service reps to focus on higher-value, more complex requests. For example, if a customer asks for a historical record of their investing activity over the course of the year, they could interact with a custom-built assistant (think their own exclusive ChatGPT) – and voila, that statement appears without the customer service rep having to get involved at all. As Bain notes in a recent blog, “Generative AI can be used to address the very issues that traditional AI struggles with, specifically having an answer for all questions, responding to never-before-seen problems and exercising judgment. In other words, generative AI provides an interface to simplify complexity.”
For fintechs and advisors who want to scale their businesses while delivering excellent client service, the ability to create these LLMs using custody/clearing firms’ data is a major game-changer.
A new level of intelligence
Data is power. Because custodians are in a pole position as data aggregators, now custodians and clients can collaborate to imagine new ways in which this data could be used.
This new level of intelligence could be used to identify and fight fraud. AI has already made significant inroads in helping firms detect and fight fraud – but generative AI will strengthen fraud detection to new levels. Previous AI and machine learning based models still needed to be enhanced by humans – updated with new parameters and improving upon themselves within predefined characteristics.
Generative AI, however, through its powerful natural language processing capabilities, can ingest more sources of information and teach itself how to constantly update its fraud models. These models could be reading fraud publications, taking in data from the Secret Service or other regulatory bodies, incorporating real-time fraud alerts, and constantly improving models, without human interaction.
In an environment where fraud detection is escalating in importance, this ability could be game-changing for firms who need to better identify suspicious activity that humans simply can’t see or detect.
Generative AI-driven data intelligence could also be used to allow custodians holding client cash to drive more optimal investments of client cash, whether altering the duration of treasuries or the mix of banks or allocation among various banks of that client cash.
Generative AI will revolutionise how both custodians and fintechs/advisors can use data in new ways to tackle numerous challenges – of today and tomorrow.
Easier, more intuitive customer account openings
For fintechs and advisors who want to onboard customers who might be transferring from another firm, this process is still oftentimes time-consuming and form-driven. Generative AI can intuitively help customers ensure they’ve filled out the right forms and with complete data. Imagine if a customer is transferring an account from another brokerage, but mistakenly leaves a field blank. A Generative Ai driven assistant could text or email you to say conversationally, “I can populate that form for you. Do you want me to take a look in your email to find that info? I could also email your advisor to find out.”
The speed at which Automated Customer Account Transfers (ACATs) could happen could increase exponentially, while freeing up firms’ internal resources to focus on higher value tasks, not filtering through form errors.
Accelerating to T+0
The vision to reduce settlement times down to T+0 could finally be here – and generative AI will help us get to that vision faster. T+0 hinges on real-time information about trades and counterparties. The issue has been that a party needs to process that info in real-time. Generative AI could take instructions from two counterparties, reconcile, and settle that trade instantaneously.
Every potential time delay in the current settlement process – from those analyzing the information, running the checks, ensuring funds and securities are available, subsequent report filings – these are all areas that Generative AI could ultimately assist with or accelerate, with minimal human oversight.
It’s not sci-fi — it’s here
There is a reality that all fintechs and advisors must confront — generative AI is no longer science fiction. It’s here. And those firms that start to invest in building LLMs and how to harness the power of custody and clearing firms’ data to do so will gain the competitive upper hand in the years to come.