Snowflake Summit 2025: How AI Is Reshaping the Financial Landscape

1. Automated Regulatory Reporting at Scale

Durgesh Das, vice president of data, analytics and governance at the Intercontinental Exchange (ICE), says that his team has used Snowflake’s Cortex AI to handle complex regulatory reporting requirements related to the Payment Card Industry Data Security Standard, Markets in Financial Instruments Directive II, and the U.S. Securities and Exchange Commission’s Rule 613. 

By shifting to Snowflake, ICE achieved a 50% cost reduction in reporting workloads and 80% better ad hoc query performance. With billions of transactions made every day, it’s helped them improve market surveillance and customer reporting.

RELATED: Improve your daily operations with data-driven analytics and services.

“Between all of our systems, we’ve handled more than a half trillion messages on a single trading day,” says Anand Pradhan, senior director of regulatory and NMS tech at the New York Stock Exchange, in this Snowflake case study. “All these transactions happen in microsecond granularity, so that produces complex, dense time-series data.”

“From our perspective, we have two incredibly important jobs. One is around market integrity, focus on transparency. The other is around efficient risk attitude. So, ensuring that those messages get processed incredibly efficiently is key,” said Lynn Martin, president of NYSE Group in a keynote conversation with Snowflake CEO Sridhar Ramaswamy. 

“We can’t do that without having incredible technology. AI enables us to match those trades really quickly,” Martin said.

2. Granular Data Lineage and Governance for AI Pipelines

Financial services are also using AI to create traceable, well-governed data pipelines that are especially helpful when dealing with sensitive trading or customer data.  

Snowflake CISO Brad Jones said that it all comes down to “having that underlying data classification and governance in place.”

“If you don't start with that solid foundation, everything falls apart from there. Do it at the point of data ingestion, initial analysis and when you’re working with semantic models,” he said.

Jones said that features such as Snowflake’s built-in data masking and row-level access policies are making it easier to protect personally identifiable information in analytics workflows; maintain compliance with regulations; and ensure that data scientists work with governed, de-risked data sets.

KEEP READING: Snowflake Summit experts share how to make data and AI more trusted.

But even so, AI requires oversight. “You're always going to have to have a human in the loop to monitor what’s okay to do with data and what's not okay to do with it,” he said.

CDW experts call this a “minimum viable data governance” strategy. It includes “security, privacy and meta-data management,” said Ben Castleton, principal consultant for data quality at CDW. Even something as simple as “keeping track of metadata, what the column names are, what the synonyms are for and how they interact with each other” will improve the quality of analytics, he said: “LLMs can't quite figure that out, but once you give it that sort of information, you're set up for success.”

3. Using Agents To Meet KYC and AML Requirements

Banks are also turning to AI agents to meet the stringent standards of a process called know your customer (KYC), as well as anti-money laundering (AML) processes. These standards are difficult to meet because they require continuous monitoring and the rapid analysis of vast and varied data sources — all without disrupting customer experience.

HSBC integrated Google Cloud’s AML agent, resulting in a 60% reduction in false alerts. “We check about 1.35 billion transactions for signs of financial crime each month, across 40 million customer accounts. We’re using AI to help us do this,” writes Jennifer Calvery, group head of financial crime at HSBC, in a company blog post.  

Valley Bank, located in Wayne, N.J., implemented an AI agent named Tara that runs automated transaction alerts for sanctions. This helped the bank achieve a 65% automation rate and improve processing efficiency of over 20,000 alerts per month.

These agents are performing a variety of tasks, including scanning structured and unstructured documents (invoices, email logs) for KYC verification; detecting suspicious behavior by correlating across customer account activity, location and transaction history; and auto-generating alerts and compliance summaries for human review.

FIND OUT: CDW can help financial services meet complex compliance regulations.

4. Embedding Risk Models into Decision Models and Audits

Banks are also embedding pretrained models (via Snowpark and Cortex AI) into credit and fraud risk pipelines to run on-the-fly scoring and evaluate payment transactions. Snowflake’s data platform also includes new citation features that show where data comes from.

These simplified reporting systems helped FIS, a financial services company in Jacksonville, Fla., improve internal audits. 

“After consolidating the work of three separate systems (Sybase, SQL Server and Hadoop) into one single source of truth in Snowflake, FIS now processes compliance data up to 20 times faster,” writes Patrick Donohue, senior director for enterprise strategy at FIS, in this Snowflake case study.

UP NEXT: The benefits of building an AI center of excellence.

Experts at the summit said that Snowflake Horizon and Cortex AI are helping banks track model versioning, data lineage and access history. This helps teams feel confident that every AI decision is reproducible, transparent and defensible to regulators.

“The data volumes and processing demand for these compliance reviews and reports can be rather intense,” noted Nathan Call, director of RegTech solutions at FIS, in the case study. “So, the technology we use to build the compliance systems can really make or break them.”

Keep this page bookmarked for articles from the event, follow us on X (formerly Twitter) @BizTechMagazine and join the event conversation at #SnowflakeSummit.