BOSTON–As digital payments and real-time transactions expand, fraud detection and anti-money laundering (AML) systems are generating more alerts than compliance teams can investigate, creating mounting backlogs across financial institutions, according to a new analysis from PYMNTS Intelligence.
The surge in alerts has created operational pressure for banks, fintech firms and payment providers that must meet strict regulatory expectations while handling a rapidly rising volume of suspicious-activity notifications, PYMNTS reported.
“Fintechs, banks and payment providers today are dealing with alert volumes that grow faster than their teams can handle,” Madhu Nadig, co-founder and chief technology officer of compliance technology firm Flagright, told PYMNTS. “Every screening hit, every transaction monitoring hit needs to be investigated, even though a significant portion turns out to be nothing. Resource requirements rise faster than teams can scale.”
‘Alert Overload’
The phenomenon, often called “alert overload,” has become one of the most persistent bottlenecks in financial crime compliance, according to PYMNTS Intelligence.
AI tools entering the investigation process
To address the growing workload, a new category of artificial intelligence tools — described as “AI forensics” — is emerging to support compliance teams, PYMNTS said, adding that these systems rely on specialized AI agents designed to perform specific investigative tasks within AML compliance and fraud prevention.
‘Specialized Agents’
“AI forensics is a family of specialized AI agents, each purpose-built to perform a specific investigator task across AML compliance and fraud prevention,” Nadig said, according to the report. “You can think of them as digital investigators that follow your institution’s standard operating procedures exactly the same way your analysts follow them, but they can execute them autonomously at scale and in a few seconds.”
The report explained that under this model, traditional rules-based monitoring systems continue to detect suspicious activity. The AI layer then performs investigative work, gathering data, summarizing evidence and preparing case files for analysts.
In some cases, AI agents can also autonomously investigate and close low-risk alerts, allowing compliance staff to focus on more complex cases.
Scaling the Process
The report notes that compliance teams often face a mismatch between alert volumes and investigative capacity. For example, a team might be able to review roughly 1,000 alerts in a week while systems generate several thousand during the same period, immediately creating a backlog.
“A human analyst cannot scale beyond a certain point,” Nadig told PYMNTS. “They need that time to make the judgment call.”
AI-driven investigative tools can reduce the time required to review each alert by automatically collecting information from multiple systems and preparing summaries for analysts, the report added.
“AI agents can pre-investigate alerts for you and bring the average investigation time from something like five minutes to maybe a minute,” Nadig said.
