
The call came in at 2:40 on a Tuesday afternoon, and the number on the screen matched the one printed on the back of the debit card. The man gave a badge number, said accounts had been opened that morning in three states, and said the only way to stop the next withdrawal was to move the balance into a holding account. He read out a routing number twice and waited while it was written down. The wire left at 3:15 p.m. for $18,400, the entire savings balance. At the branch the next morning, the manager pulled the record, looked up, and said the transfer was authorized.
That answer is correct in the only sense the governing rule cares about. Federal protection for electronic transfers out of a consumer account turns on six words buried in one definition: a person other than the consumer. 12 CFR 1005.2(m) defines an unauthorized electronic fund transfer as one “initiated by a person other than the consumer without actual authority to initiate the transfer and from which the consumer receives no benefit.” A transfer the consumer typed, confirmed, and sent fails that first element before any question of deception arises.
There is a version of the same afternoon that ends differently. If the caller had talked the account holder into reading out a login and a texted confirmation code, and had then moved the money himself, the Consumer Financial Protection Bureau’s own compliance guidance says the result is an unauthorized transfer with full liability protection. Nothing about the lie changes between the two versions. What changes is whose hands were on the keyboard when the money moved.
The line between a covered claim and a denied one is drawn by who pressed send, not by how convincing the story was. A recall is a request the receiving institution is free to refuse. And the machinery that can freeze the funds sits at the FBI and the Treasury Department rather than at the bank, runs on a clock measured in hours, and has to be started by the person who sent the money.
The six words that decide the claim
Regulation E, which implements the Electronic Fund Transfer Act, carries its entire fraud allocation in the definition at 1005.2(m). The positive test requires a transfer “initiated by a person other than the consumer,” made “without actual authority,” and “from which the consumer receives no benefit.” The elements are cumulative, which is why the third is no help here: a transfer that fails the first never reaches it.
Three carve-outs follow, excluding a transfer initiated:
- “By a person who was furnished the access device to the consumer’s account by the consumer, unless the consumer has notified the financial institution that transfers by that person are no longer authorized.”
- “With fraudulent intent by the consumer or any person acting in concert with the consumer.”
- “By the financial institution or its employee.”
The official interpretations work the edges, and two of the five comments do most of the work. Comment 2(m)-3 reads: “an unauthorized EFT includes a transfer initiated by a person who obtained the access device from the consumer through fraud or robbery.” Under Comment 2(m)-4, “an EFT at an ATM is an unauthorized transfer if the consumer has been induced by force to initiate the transfer.”
Comment 4 is the one place in the commentary where a consumer’s own hands on the terminal still produce an unauthorized transfer, and the trigger there is force, not persuasion. A person marched to an ATM at knifepoint is covered. A person talked into a wire by a stranger claiming to be from the fraud department is not addressed at all, because the drafters treated a transfer the consumer chose to make as the consumer’s transfer.
Comment 2(m)-2 handles the case that looks closest to a scam and is treated furthest from one: a consumer who furnishes an access device to someone “who exceeds the authority given” is “fully liable for the transfers” until the institution is told that person is no longer authorized. Handing the card and PIN to a relative who then takes more than agreed is a dispute between two people, one of whom has the money.
Credentials taken, or a payment sent
The Bureau’s Electronic Fund Transfers FAQs, a compliance aid at Version 3 and last updated January 15, 2025, spends a full section on unauthorized transfers.
Question 5 asks whether the definition is met when “a third party fraudulently induces a consumer into sharing account access information that is used to initiate an EFT from the consumer’s account.” The answer is yes. One of the Bureau’s examples is “a third-party calling the consumer and pretending to be a representative from the consumer’s financial institution and then tricking the consumer into providing their account login information, texted account confirmation code, debit card number, or other information that could be used to initiate an EFT out of the consumer’s account.”
Question 6 closes the escape route institutions reach for first. Asked whether transfers made with fraudulently obtained credentials fall inside the first carve-out because the consumer handed the information over, the Bureau answers no: “A consumer who is fraudulently induced into providing account information has not furnished an access device under Regulation E.”
The rest of the section removes the usual defenses. Question 7 says an institution may not weigh the consumer’s carelessness, quoting Comment 6(b)-2 that writing a PIN on the card “does not affect the consumer’s liability for unauthorized transfers under Regulation E.” Question 8 forecloses contracting around any of it, citing the anti-waiver provision at 15 U.S.C. 1693l. Nor can network rules do that work. Under Question 10, finality and irrevocability between banks “do not reduce consumer protections against liability for unauthorized EFTs afforded by the Electronic Fund Transfer Act.”
That is a strong set of protections, and every one describes the same fact pattern. In Question 4’s examples a hacker breaks into a merchant’s systems, a fraudster hacks a phone and uses the mobile wallet, a thief steals a wallet; in Question 5 the third party takes the login and uses it. The fraudster initiates. The section contains no question in which the consumer opens the app, enters an amount, and taps send, because that transaction is not an unauthorized transfer.
The payments the rule never reached
For a wire, the threshold question is not liability but coverage. 12 CFR 1005.3(c)(3) excludes from coverage “any transfer of funds through Fedwire or through a similar wire transfer system that is used primarily for transfers between financial institutions or between businesses.” The commentary names the similar systems: “the Clearing House Interbank Payments System (CHIPS), Society for Worldwide Interbank Financial Telecommunication (SWIFT), Telex, and transfers made on the books of correspondent banks.”
A retail wire from a checking account runs on those rails, and the exclusion does not turn on who initiated it. The commentary points instead to state law, citing UCC section 4A-202 for a telephone-initiated Fedwire payment. UCC 4A-108 states the division from the other side: “this Article does not apply to a funds transfer any part of which is governed by the Electronic Fund Transfer Act of 1978.” For a consumer wire, then, the governing law is Article 4A, a commercial code written for banks and businesses rather than a consumer protection statute.
One carve-in matters. Subpart B covers remittance transfers to recipients abroad, with 1005.30(e) excluding transfers of “$15 or less” and 1005.30(f) providing a safe harbor to a person that both provided and provides “500 or fewer remittance transfers” in the previous and the current calendar year. What it adds is a clock most senders never hear about. 12 CFR 1005.34 requires a provider to honor a cancellation request received “no later than 30 minutes after the sender makes payment,” so long as it identifies the sender and the particular transfer and “the transferred funds have not been picked up by the designated recipient or deposited into an account of the designated recipient.” The refund then covers the total amount provided, fees and taxes included, within three business days. Thirty minutes is short and that second condition is the one that usually fails, but for a transfer where doubt sets in on the walk back to the car, the deadline is worth knowing.
Whether the institution behind an app-based transfer is even a bank is a separate question worth settling first, and the answer is not always what the marketing implies, as When FDIC Insured Means a Partner Bank, Not Your App works through.
What a recall request can and cannot do
The word the branch will use is recall, and it sounds like a reversal. Article 4A defines it narrowly. UCC 4A-211 lets a sender cancel a payment order “orally, electronically, or in writing,” effective under subsection (b) only if it arrives in time for the receiving bank “to act on the communication before the bank accepts the payment order.” That window closes in the minutes before acceptance, not the hours after a scam is discovered.
Subsection (c) then decides most cases: “After a payment order has been accepted, cancellation or amendment of the order is not effective unless the receiving bank agrees or a funds-transfer system rule allows cancellation or amendment without agreement of the bank.” Once the beneficiary’s bank has accepted, the sending bank has no power to claw the payment back. It can ask. The receiving institution decides, weighing its own exposure to a customer whose account it would be debiting.
Hence the indemnity attached to the recall paperwork. The FBI’s Internet Crime Complaint Center states the first move in its published domestic kill chain process: “Contact the originating Financial Institution as soon as fraud is recognized to request a recall or reversal as well as a Hold Harmless Letter or Letter of Indemnity.”
The Office of the Comptroller of the Currency sets the ceiling on expectations. Asked whether a customer can be held liable after wiring money to a scammer, HelpWithMyBank.gov answers: “Your bank would need to review your fraud claim. Wire transfers are intended to be a final and irrevocable method of payment and the bank may not be able to reverse the wire transfer.” Its page on what to do about a fraudulent wire adds the instruction easiest to miss, which is to call both ends: “Contact your bank and the bank that may have received your funds.”
The ACH side is narrower still. Nacha’s rules permit a reversing entry only to correct an erroneous entry, and Nacha’s own summary lists what qualifies: a “duplicate payment, incorrect amount, payment sent on the wrong date, or payment sent to an unintended account,” transmitted “within five banking days of the original settlement date.” A payment the sender meant to send, in the amount intended, to the account number written on a receipt, is not on that list. What the bank sends instead is a request, answered at the receiving institution’s discretion.
Underneath all of it sits a timing problem no letter solves. Accounts opened to receive scam proceeds do not hold the money long, which is the premise both federal programs are built on. It is also why receiving institutions close them, and why the people who opened them often cannot open another, a consequence covered in Denied a Bank Account Because of ChexSystems: What to Do. By the time a recall request is drafted, the question is usually not whether the receiving bank will agree but whether anything is left to agree about.
The first seventy-two hours
Two federal programs aim at this gap.
The Financial Crimes Enforcement Network’s Rapid Response Program, described in a fact sheet issued February 11, 2022, states the finding directly: “the RRP has had greater success in recovering funds when victims or financial institutions report fraudulently induced wire transfers to law enforcement within 72 hours of the transaction.” The fact sheet puts the cumulative result at “more than $1.1 billion” recovered for U.S. victims since 2015, working through counterpart financial intelligence units abroad.
The activation path is specific: “A victim of a cyber-enabled crime, or the victim’s financial institution, must file a complaint with law enforcement to initiate the RRP,” through the FBI’s complaint center or the nearest Secret Service field office, and “Victims should not directly contact FinCEN.” The required fields are worth writing down before calling, because a complaint missing any of them stalls:
- Victim’s name, the victim’s financial institution’s name, and the country location of the originating branch
- The victim’s account name, which the fact sheet notes “may be different than the victim’s name,” and account number
- The beneficiary’s financial institution’s name and the country location of the receiving branch
- The beneficiary’s account name and account number
- Date of the wire transaction, and the currency and amount transferred
The fact sheet closes that list with one line: “Time is of the essence to successfully recover funds!”
The domestic counterpart is the Recovery Asset Team at the FBI’s Internet Crime Complaint Center, established in February 2018, which runs the Financial Fraud Kill Chain. The 2025 IC3 Annual Report gives the year’s totals: 3,900 incidents, $1,163,919,846 in attempted theft, $679,013,183 frozen, a 58% success rate. It notes a shift away from business email compromise toward tech support and account takeover cases, the latter sometimes involving “50 or more transactions to different recipient accounts at multiple banks happening simultaneously via ACH transactions.”
Its guidance to victims repeats both halves of the sequence: “If you discover a fraudulent transfer, time is of the essence. Immediately, contact your financial institution and request a recall of the funds along with any necessary indemnification documents.” Then: “Regardless of the amount lost, file a complaint at www.ic3.gov.” That second instruction is unconditional, and the report’s note that “different financial institutions have varying policies” warns that the first step may produce nothing.
Against $20.877 billion in total reported losses for 2025, a 58% freeze rate on the cases that reach the team in time is the best number in this subject, and the one that decays fastest with delay.
Where each report actually lands
The reports people are told to file after a scam do different jobs.
A complaint at ic3.gov is the one with a recovery mechanism bolted to it. It triggers the Recovery Asset Team and is the route the FinCEN fact sheet names for reaching the international program. Transaction detail is not a formality; the freeze request sent to the recipient bank is built from the account numbers in the complaint.
ReportFraud.ftc.gov feeds the Federal Trade Commission’s enforcement and data work. The agency’s guidance on what to do if you were scammed is organized by payment method and, in every case, points first at the company that moved the money rather than at itself: report a wire “to the wire transfer company immediately,” a payment app payment “to the payment app immediately.” That ordering describes where the leverage sits.
The Consumer Financial Protection Bureau’s complaint process is a routing service with a clock. The Bureau forwards it to the company, and “companies generally respond in 15 days,” with some taking up to 60. It does not adjudicate, but it does put a written answer on the record from an institution that may otherwise have given only a verbal denial at a counter.
Which prudential regulator hears a complaint depends on the charter: national banks and federal savings associations answer to the OCC through HelpWithMyBank.gov, federal credit unions to the National Credit Union Administration, state-chartered institutions to a state regulator and often to the Federal Deposit Insurance Corporation or the Federal Reserve. Working out which applies is the lookup described in How to Verify a Bank Is Actually FDIC Insured Before You Transfer, run in reverse. A local police report belongs on the list for an unrelated reason: insurers, employers, and adult protective services routinely ask for a report number, and it is easier to get in week one than in month three.
None of these returns money on its own. One of them can freeze it.
The error resolution that still applies
A denial under 1005.2(m) does not empty the file. 12 CFR 1005.11 defines an “error” as a list of seven items. The first is “an unauthorized electronic fund transfer.” The seventh is “the consumer’s request for documentation required by § 1005.9 or § 1005.10(a) or for additional information or clarification concerning an electronic fund transfer.”
That seventh item is available whether or not the first is. A written request for the transfer record, the timestamps, the channel the instruction arrived through, and the institution’s own findings is itself a notice of error, and it starts the same clocks: 10 business days to investigate and three more to report, extendable to 45 days if the account is provisionally credited within the 10. The outer limit stretches to 90 days for a transfer that was not initiated within a state, a point-of-sale debit card transaction, or one that occurred within 30 days after the first deposit to the account.
Section 1005.11(d) governs the denial itself. The institution must give “a written explanation of the institution’s findings” and tell the consumer of the right to request the documents it relied on. A verbal no at a counter does not satisfy that, and those documents are the only way to test whether the institution determined who initiated the transfer or simply assumed it. Under 15 U.S.C. 1693f(e), the consumer is entitled to treble damages where a court finds the institution did not provisionally recredit within the ten-day period and either “did not make a good faith investigation” of the alleged error or “did not have a reasonable basis for believing that the consumer’s account was not in error.”
Where the facts are contested, the two regimes touch. If the claim is that a fraudster used a captured passcode, the FAQ answers put that inside the definition and the liability tiers at 1005.6 apply. Those tiers, and the 60-day statement deadline governing all of them, are worked through in The 60-Day Line Where Debit Card Fraud Stops Being Capped. Section 1005.6(b)(4) also requires an institution to extend the deadlines to “a reasonable period” where the delay “was due to extenuating circumstances,” which is the provision that matters when a hospital stay sits between the statement and the discovery.
Instruments that carry their own protection
The protection is in the instrument, and it is chosen before the money moves. A credit card carries two separate rights. 12 CFR 1026.12(b) caps liability for unauthorized use at “the lesser of $50 or the amount of money, property, labor, or services obtained by the unauthorized use before notification to the card issuer.” Independently, 12 CFR 1026.13 defines a billing error to include an extension of credit “for property or services not accepted by the consumer or the consumer’s designee, or not delivered to the consumer or the consumer’s designee as agreed.” Goods paid for and never shipped are a billing error on a credit card, and nothing comparable exists for a completed push payment out of a deposit account. Notice must be in writing and reach the creditor “no later than 60 days after the creditor transmitted the first periodic statement that reflects the alleged billing error,” and the creditor must resolve it within two complete billing cycles, “in no event later than 90 days.”
Section 1026.12(c) adds a third right with fine print. A cardholder who has tried in good faith to resolve a dispute with the merchant “may assert against the card issuer all claims (other than tort claims) and defenses arising out of the transaction,” but only where the amount “exceeds $50,” and the transaction occurred in the cardholder’s own state or “within 100 miles from that address.” Those limits fall away where the merchant is the card issuer or is controlled by it.
A debit card sits under Regulation E’s tiers, weaker than a credit card and stronger than a wire, because an unauthorized debit card transaction falls inside 1005.2(m) while a wire is outside Regulation E altogether. Checks sit outside coverage too, under 1005.3(c)(1), which leaves a forged or altered check to a separate body of state law.
Ordered by what happens when the counterparty turns out to be a stranger who lied, the sequence runs credit card, debit card, ACH, wire, then cash and its equivalents. A request to move money by a method at the far end of that list, from someone who called first, is the tell.
Verification runs in the other direction too, for money arriving rather than leaving. A deposit that posts is not a deposit that is final, and the gap between the two is the mechanism behind overpayment scams, which is why Why Your Deposit Is Not Available Yet: Funds Availability Rules is worth reading before refunding anyone a difference.
Common questions
The caller’s number matched the bank’s exactly. Does spoofing change the analysis? Not under 1005.2(m), which turns on who initiated the transfer. Spoofed caller ID goes to how convincing the deception was, not to who entered the payment instruction. It matters for the criminal complaint: the number belongs in the IC3 filing with any text messages, which the FinCEN fact sheet lists among the secondary information useful for tracing.
Is a wire really final as a matter of law, or is that the bank’s policy? Closer to law than policy. UCC 4A-211(c) provides that after a payment order has been accepted, “cancellation or amendment of the order is not effective unless the receiving bank agrees.” The OCC’s consumer site describes wires as “intended to be a final and irrevocable method of payment.” A sending bank can request a recall; it cannot compel one.
Does filing with several agencies improve the odds? They are not interchangeable. The complaint at ic3.gov is the one connected to the Recovery Asset Team and, through it, to the FinCEN program for funds sent abroad, and the fact sheet is explicit that victims should not contact FinCEN directly. A CFPB complaint produces a written company response on a 15-day clock; an FTC report feeds enforcement data. The IC3 complaint, with complete account numbers on both sides, is the one under time pressure.
A scammer got a one-time passcode and moved the money. Is that also outside the rule? No, and the distinction is the whole point. FAQ Question 5 treats a third party who tricks a consumer into providing a “texted account confirmation code” and then initiates the transfer as having made an unauthorized transfer, and Question 6 confirms the consumer has not “furnished an access device” by being deceived. That claim runs through 1005.11 error resolution and the 1005.6 liability tiers.
The bank denied the claim over the phone. Is anything owed in writing? If a notice of error was given, yes. Section 1005.11(d)(1) requires a written explanation of the findings and notice of the right to request the documents the institution relied on. A request for documentation or “additional information or clarification concerning an electronic fund transfer” is itself an error under 1005.11(a)(1)(vii), so asking in writing starts the clock even where the underlying transfer is not covered.
The rule was written in the 1970s around a straightforward question about a card and a terminal: did the account holder do this, or did someone else. A payment instruction a stranger can talk a person into issuing from a phone in a parking lot answers that question cleanly, in the direction that ends the claim. The branch manager on Wednesday morning was reading the definition correctly.
Further Reading
- Why You Owe Tax on a 5-Year CD You Cannot Touch Yet
- I Bonds vs a High-Yield Savings Account: The 12-Month Lock
- Treasury Bills vs CDs: The Tax Break That Changes the Comparison
- CD Early Withdrawal Penalties: The Disclosure and the Deduction
- Backup Withholding on Savings Interest: Why the Bank Took 24%