Smarsh Launches AI-Enabled Communication Surveillance for Small to Medium Businesses

WHY THIS MATTERS:
Financial institutions are facing a surge in digital communications across email, messaging platforms and collaboration tools, while regulators continue to tighten oversight of recordkeeping and conduct monitoring. For many compliance teams — particularly at small and mid-sized firms — the challenge is not detecting misconduct but managing the overwhelming volume of alerts generated by surveillance systems. Smarsh’s new Noise Reduction Agent aims to address this by using AI to suppress low-risk communications during the ingestion stage, before alerts reach compliance review queues.

By filtering out spam, disclaimers, automated system messages and other non-actionable content while maintaining full audit records, the system reportedly reduces false positives by up to 60% and saves significant reviewer time each month. This reflects a broader shift in compliance technology toward AI-driven automation that prioritizes signal over noise, allowing human reviewers to focus on genuine regulatory risks rather than processing large volumes of irrelevant alerts.

As compliance teams across financial services struggle to keep pace with exploding volumes of digital communications and intensifying regulatory scrutiny, Smarsh today announced the launch of its Noise Reduction Agent, an autonomous AI capability designed to dramatically reduce compliance alert volume up to 60% and eliminate thousands of hours of non-actionable review work each month.  

For small and mid-sized financial firms operating under the same regulatory obligations as the world’s largest banks—but without comparable staffing—alert fatigue has become a structural risk. Reviewers often process thousands of daily communications, the majority of which pose no regulatory concern.

“Compliance teams aren’t [MS1] overwhelmed by misconduct—they’re overwhelmed by the volume of noise,” said Sheldon Cummings, President, Corporate Business at Smarsh. “Noise Reduction Agent allows firms to suppress low-risk communications before they ever reach supervision queues, dramatically reducing review backlogs while preserving full regulatory defensibility.”

Early preview reports have shown promising results.

•   Up to 60%+ reduction in false positives
•   40+ hours saved per reviewer per month
•   20% increase in redirected capacity

How it Works

Today, nearly half of all global email traffic is spam, highlighting the scale of the noise compliance teams must filter every day. Unlike traditional filtering tools that operate after alerts are generated, Noise Reduction Agent applies AI during ingestion within Smarsh Professional Archive. The system automatically identifies and suppresses low-risk email content—such as spam, disclaimers, newsletters, promotional text, and automated system messages—while preserving complete records for audit and regulatory purposes.

This proactive approach enables firms to reduce operational burden without compromising supervision, compliance, or defensibility and allowing them to focus on real compliance issues.

Expanding AI Beyond Global Banks

While AI-powered surveillance has historically been deployed primarily by the largest financial institutions, Smarsh is making mature AI capabilities accessible to smaller compliance teams for the first time.[AB2] 

The Noise Reduction Agent builds on more than a decade of AI innovation deployed across the world’s largest financial institutions and reflects a broader shift toward autonomous AI systems that augment human expertise rather than replace it. 

“We’re proud to deliver a scalable surveillance solution grounded in best-in-class AI governance and model risk management,” said Maya Keekattil, VP of Product Management. “Our mission has been clear: bring enterprise-grade AI supervision to small and mid-sized firms in a way that is accessible, affordable, and built for regulatory confidence.”

FF NEWS TAKE:
Compliance technology is entering a new phase where AI is applied to operational efficiency, not just detection.

Alert fatigue has long been a structural problem in financial supervision. If tools like Smarsh’s Noise Reduction Agent can reliably reduce false positives while maintaining regulatory defensibility, they could significantly reshape how compliance teams manage surveillance workloads — particularly for smaller firms operating under the same regulatory expectations as global banks.