COFFR · THE PAYMENTS TRUST LAYER

$340,000Gone in the time it took to hit send.

Somewhere today, a finance team is wiring money to an account that looks exactly right. It won't be. This is the story that keeps repeating, and it's why we built the layer that stops it from happening at your company.

This isn't about careless companies. It's about good people, at good companies, doing exactly what they were trained to do.

01Cited, primary-source research
02Case patterns from the public record
03Weekly industry signals

The impact of ACH/Wire fraud

Anatomy of
BEC Fraud
ACH and wire, at scale
$3.046B
Reported BEC losses in the latest IC3 report
FBI IC3 2024
24,768
Verified BEC complaints filed with federal law enforcement
FBI IC3 2024
$120K+
Average loss per BEC incident
FBI IC3 2024
74%
Of fraud attempts hit ACH first
AFP Payments Fraud

BEC losses rose from $2.77B to $3.046B in the most recent IC3 report. Generative AI is projected to accelerate fraud losses roughly 32% a year through 2030.

Your controls did not fail. They worked exactly as designed.

That is the uncomfortable part. Coffr is an independent research effort focused on one narrow question: how do two companies prove they know who they are paying, before the money leaves the account? Every piece we publish sits under one of these three threads.

The vector

They were already inside the email thread

Thread hijacking, lookalike domains, forged remittance PDFs, and AI-assisted impersonation. We track how the initial contact actually happens, case by case.

From this threadAI agent payment controls should require a close file that proves the named customer authorized the specific invoice before release6 min read · August 11, 2026

The process

The five-minute step that costs six figures

Vendor onboarding, change-of-bank-detail requests, and approval chains. Most losses trace to a routine step that was never designed to be adversarial.

From this threadPayment vendor acquisition controls still trust outdated tools.5 min read · August 11, 2026

The record

What regulators know that your board does not

FBI IC3, FinCEN advisories, AFP surveys, Nacha rule changes, and court filings. We read the primary record and summarize what changed.

From this threadYou approved the PDF. Nobody verified the bank account7 min read · June 14, 2026

You get the email Thursday afternoon. New account, usual vendor, same tone as always. Nothing about it looks wrong, because nothing about it was built to look wrong.

You send it.

Monday, the real vendor calls asking where their money is.

This is the actual shape of almost every case we've studied. Not a hack. Not a typo. A document that had no business being trusted, and was.

One email a week. The case nobody wanted public.

One email each Thursday. No spam, one-click unsubscribe.

Sector register

If your company pays anyone, you are already on this list.

The exchange layer looks the same across industries. The stakes don't. These are the sectors our research covers most closely.

08 sectors under watch

01

Community & Regional Banks

Wire desks, treasury ops

02

Fintech & Neobanks

Onboarding, money movement

03

Accounting & Bookkeeping

Vendor setup, client trust

04

Law Firms

Trust accounts, client funds

05

Real Estate & Title

Closings, escrow, wire

06

Construction & Trades

Subcontractor payouts

07

Healthcare Payers

Provider disbursements

08

SMB Finance Teams

AP, vendor onboarding

Every wire fraud case we've read reduces to the same sentence: someone trusted a document that arrived by email, and no system on either side was in a position to disagree.

This isn't about careless companies. It's about good people, at good companies, doing exactly what they were trained to do.

Field notesThe Coffr research deskDrawn from published wire fraud filings and settlement records
07 / Signals

The headlines your peers are quietly forwarding.

A curated feed of banking, fraud, and payments coverage from regulators and industry press, updated continuously.

All signals

Feed temporarily unavailable

We'll be back with the latest coverage shortly.

ii / PanelBy request

You have seen this go wrong. Tell us where.

Practitioners on the research panel get short surveys about payment verification and fraud controls, plus the findings before they are published.

This is research participation. It is not early access, and it is not a product waitlist.

Nobody tells you how payment trust actually fails.

We publish what we learn: the vectors, the case patterns, the regulatory record, and the industry signals worth tracking. Get the weekly briefing, or join the research panel.