Employees know AI is making fraud harder to catch, but they’re not even looking for the right signals. Trustmi surveyed 1,000 U.S. employees who work with invoices, vendor communications, payment approvals, and finance systems, and discovered a problem.
While 78% say they’d recognize fraud, 70% are still watching for the typos and bad grammar of yesterday’s phishing. Today’s AI-driven fraud arrives polished, inside the email threads, names, and projects they already trust.
This report breaks down where employee confidence and real-world defenses diverge: the warning signs people still rely on, the trusted signals attackers exploit, and why 79% of employees say email filters alone aren’t enough to stop modern payment fraud.
A quick, sharable read for security and finance leaders rethinking how their organizations verify payments.

Behavioral AI-powered security
Protection on day one
10-15x ROI