How can a company spot sanctions violations and financial crime? Through AI

Windward CEO Ami Daniel explains why preventative policies are not enough to deter misbehavior.

This is every risk manager’s nightmare: Despite clear compliance rules in place, rogue employees acting in their self-interest knowingly violate internal regulations – exposing the company to hefty fines, or worse.

That’s what happened at Japanese trading firm Sojitz Corp. after several errant employees at its Hong Kong subsidiary hid a trade deal’s Iran connection, according to the U.S. Treasury Department’s Office of Foreign Assets Control. The dollar-denominated transaction was processed through U.S. financial institutions, causing them to violate U.S. Iran sanctions regulations. Sojitz voluntarily disclosed this internal breach to authorities and was fined $5.2 million.

Artificial intelligence can help companies identify sanctions violations and financial crime in maritime trade even if there is rogue behavior among employees, said Ami Daniel, CEO and co-founder of Windward, a maritime data and analytics company, in an interview with AI Business. His firm analyzes trade data of voyages, ports, vessels and other shipment entities.

Preventative policies are not enough

 A company like Sojitz that engages in offshore trading and cross-border trade financing would have such policies around sanctions compliance, but those policies failed to detect the activities of rogue employees.

“Sanctions compliance is not just having a policy, but also ensuring that the procedures are enforced,” Daniel said. "This is why AI solutions are so critical -- having a policy is not enough as it's susceptible to human error or human manipulation.”

He explained that AI systems could track containers or ships and flag any activity seen as suspicious -- such as visiting a port in a sanctioned country, turning off their tracking near sanctioned ports, and using other deceptive shipping practices.

Bad actor prevention in finance

As criminals deploy more sophisticated methods to evade compliance teams, AI systems become a critical defensive tool to find patterns of misbehavior. For example, financial firms use AI for risk-profiling to vet clients against current sanctions, PEPs (politically exposed persons), enforcement lists and adverse or negative media.

“Limiting ongoing screening to those customers who pose the greatest risk helps banks’ balance risk versus resources,” Daniel added. “AI solutions are powerful tools which save precious time and can alert firms about risks that a human can easily miss.”

“However, there are ethical dilemmas involved with allowing AI to completely run operations and humans must be the ones to make the final go/no-go decision on accounts with AI empowering them to make these decisions better and faster,” he said.

Looking ahead, behavioral analytics tools that scrutinize the actions of users are going to become more critical in identifying risk patterns, Daniel said.

“Financial institutions must go beyond their walls to identify how their customers are doing business. Sanctions compliance requirements will continue to grow and tools like Windward will enable banks to stay ahead of the curve,” he added.