The behavioral AI layer that verifies every business payment - before the money moves.
Trustmi's wordmark, rendered in yellow on navy. The New York- and Israel-based company builds "digital fingerprints" of vendors to catch fraud that looks entirely legitimate. - Company profile, July 2026
Every large company runs on payments it barely watches. Invoices arrive, vendors send updated banking details, approvals get rubber-stamped, and money leaves - thousands of times a month. Most of it is legitimate. Some of it is not. And the fraudulent slice increasingly looks identical to the rest.
Trustmi, founded in 2021 by Shai Gabay and Eli Ben Nun, is built for exactly that blind spot. It is a behavioral AI platform for payment security that sits across the payment lifecycle and answers one deceptively simple question about every transaction: safe, or unsafe?
Rather than scanning for malware or obvious red flags, Trustmi studies how each vendor has actually behaved over time - payment sizes, timing, banking details, communication patterns - and turns that history into a "digital fingerprint." When a new payment deviates from the pattern, the system flags it before funds leave the building.
The company says it protects more than $240 billion in payments annually and has helped prevent over $1 billion in fraud and $5 billion in payment errors. It counts Colgate-Palmolive, Monday.com and CNA Insurance among its customers, and in 2026 became a Nacha Preferred Partner for fraud prevention.
The old fraud playbook - spot the typo, catch the bad link - is retiring. AI now writes clean phishing emails, clones domains, and deepfakes the voices that approve wire transfers. Trustmi's own 2026 benchmark report put it bluntly: fraud now often arrives "pre-approved," waved through by the very processes meant to stop it.
The company's research also surfaced a structural weakness: roughly one-third of organizations said a disconnect between their finance and security teams contributed to fraud. The people who move the money and the people who guard the doors often don't share the same view of a payment. Trustmi is designed to live in that gap - unifying signals across email, ERP and financial systems that would otherwise sit in separate silos.
Its customers are enterprise finance, accounts payable and security teams across manufacturing, pharmaceutical, healthcare and insurance - industries where a single spoofed invoice can move six or seven figures to the wrong account.
Trustmi assesses historical payments to build a "digital fingerprint" for each vendor across hundreds of data points.
It links vendor, employee, communications and payment data in real time across email, ERP and financial systems.
Bank account validation uses three-way matching of company ID, bank account and ownership at onboarding and before payment.
Every payment gets a clear, real-time SAFE / UNSAFE decision - before the money moves.
Many businesses still rely on manual processes to validate payments, which can be time-consuming and prone to errors.
Analyzes activity across users, vendors, payments and systems to detect anomalies that traditional controls and manual reviews miss.
Automates onboarding, data validation and bank account verification with three-way matching to reduce fraud risk.
Unifies signals across every system to return a clear, real-time safe-or-unsafe call on each payment.
New AI agents eliminate manual review; a marketplace lets customers deploy custom agents for their own workflows.
Legacy tools ask whether a single transaction looks risky. Trustmi asks whether a payment matches how a vendor has always behaved - a shift that catches socially engineered attacks designed to pass surface-level checks. The illustrative comparison below reflects Trustmi's stated approach across the payment lifecycle.
Trustmi sells enterprise B2B SaaS: its platform integrates with existing ERP, email and payment systems, and is also available for procurement through AWS Marketplace and partner channels including Nacha's Preferred Partner program. It sits at the intersection of two markets - fintech payments infrastructure and cybersecurity.
Its neighbors and named competitors include Alloy, Hawk AI and Cable, along with accounts-payable automation tools and, most persistently, the manual verification processes still common inside finance departments. Trustmi's argument is that verification should be its own real-time software layer rather than a human chore.
The company's roots are unmistakably in security: CEO Shai Gabay served in the Israeli intelligence corps before moving into cybersecurity and fintech, and he met co-founder Eli Ben Nun in 2018 at the security firm Cynet. That heritage shows up in the product's assumption that the attacker will look legitimate - and in the company's steady output of benchmark research.
With a Smart Agent Marketplace on the way, Trustmi is betting that fraud defense can't be one-size-fits-all: a pharmaceutical company's risks aren't a manufacturer's, and customers should be able to extend the platform to their own workflows.
Shai Gabay and Eli Ben Nun launch the company to secure the end-to-end B2B payment process.
Seed capital funds development of the behavioral AI payment platform.
Cyberstarts leads a $17M round with Zeev Ventures as Trustmi launches globally.
Vendor onboarding and validation expand; customers include Colgate-Palmolive, Monday.com and CNA.
Trustmi expands into AI agents and a Smart Agent Marketplace and becomes a Nacha Preferred Partner.
Trustmi is a behavioral AI platform for payment security that verifies business payments across their full lifecycle, detecting socially engineered fraud and payment errors before money moves.
Trustmi was founded in 2021 by Shai Gabay (CEO) and Eli Ben Nun (CTO), with offices in New York and Israel.
Trustmi raised a $17M Series A in July 2023 led by Cyberstarts with Zeev Ventures, bringing total funding to roughly $21M-$25M.
Enterprise finance and security teams, including Colgate-Palmolive, Monday.com and CNA Insurance, across manufacturing, pharma, healthcare and insurance.
Instead of only scoring transactions, Trustmi builds behavioral "digital fingerprints" of vendors and correlates email, ERP and payment signals to catch fraud that looks legitimate.