Most fraud trend coverage points at what new trends are emerging. The figures institutions reported this year point somewhere older.
According to the Federal Reserve Financial Services 2026 Risk Officer Report, 32% of surveyed institutions reported an increase in counterfeit checks, 21% reported an increase in check washing, and 18% reported an increase in payee forgery. Checks are the oldest payment instrument most institutions still handle.
The fraud most likely to cost banks and credit unions money next year is not necessarily unfamiliar. But rather, it is familiar fraud that has become easier to run at volume and harder to catch at the counter.
Three shifts are worth planning around: old fraud types arriving in far greater numbers, AI raising the quality of the approach rather than inventing a new one, and a constraint on fraud teams that has very little to do with money.
Old Fraud Types, Arriving in Greater Numbers
Check fraud is the clearest example. Nothing about the mechanics has changed. A check is intercepted from a mailbox or a collection box, the payee and amount are chemically washed off, new details are written in, and it is deposited somewhere the signer will not notice for weeks.
What has changed is how organized the front end of that process has become. Mail theft is coordinated, stolen checks are resold, and the person presenting the item at the counter often has no connection to the theft at all. They were recruited online, told the deposit was a job, and handed an item that looks ordinary.
ProSight Financial Association's 2026 State of Fraud Prevention Survey found 46% of banking professionals naming check fraud as their fastest-growing fraud method, the top answer in the set. For a teller, the practical question is not whether check fraud is rising. It is whether a specific item in front of them, from a customer who seems entirely reasonable, deserves a second look.
What AI Changed, and What It Did Not
The conversation about AI in fraud tends to focus on new attack types. The more useful framing for a 2027 training plan is that AI mostly improved the quality of approaches that already existed.
- A business email compromise attempt used to give itself away through clumsy phrasing. Now the email reads the way that vendor's emails actually read.
- A caller impersonating a customer used to fumble on details. Now the caller has a cloned voice, a plausible backstory, and enough correct information to survive the first two verification questions.
What AI has not changed is where the decision happens. Someone still approves the wire, releases the funds, or updates the account details. That decision point is a person, working under time pressure, who either recognizes the pattern or does not.
Does your current training give employees opportunities to practice for that moment?
The Constraint Is Not Budget
ProSight's 2026 survey asked what limits fraud prevention effectiveness, and the top answers were operational: alert overload at 66%, talent and staffing at 65%, and data quality at 63%. Budget constraints came in at 52%, behind all three.
Budget ranks fourth, and the three constraints above it describe the conditions the work happens in: how many alerts arrive, how many people are there to work them, and how good the data is when they do.
Two of those three point at the same place, which is the people who have to act on whatever reaches them. That is an argument for treating employee capability as a fraud control in its own right, rather than as a supporting activity around the technology.
What These Fraud Trends Mean for 2027 Training Plans
Preparing for this is less about how much training gets assigned and more about what the training is built for.
Annual awareness training was built to cover topics. What these fraud trends call for is practice at decisions: a check that looks acceptable upon first glance but is not, a caller who sounds coached, a payment instruction that changed at the last minute.
Employees who have worked through those situations in training recognize them faster when they are real. Our Frontline Fraud Prevention Conversation Guide covers what comes next, once someone decides something is wrong.
Download the 2027 Fraud Trends & Predictions Guide for a fuller picture of what to prepare for. It covers five trends heading into next year, what each one changes, and an eight-question readiness check you can run against your own program.
Prepare Your Teams for the Fraud Attempts Ahead
A few shifts in your training approach can make a significant difference:
- Scenario practice rather than topic coverage, so employees rehearse the judgment call
- Role-specific content, because a teller, a call center agent, and a wire operator face different versions of the same scheme
- Quarterly reinforcement rather than an annual assignment, so examples keep pace with what is actually being attempted
- Coverage across the institution, not only the branch, since fraud reaches contact centers and back office teams too
BankersHub's fraud prevention training helps banks and credit unions strengthen fraud prevention through scenario-based learning that is updated as tactics change, and delivered by role across frontline, contact center, operations, and fraud teams.
INTRUST Bank took this approach with recurring scenario-based training for roughly 275 frontline employees each quarter, later extending it to its customer service center. Read how INTRUST built ongoing training into its fraud reduction strategy.
