Somewhere in the United States right now, someone is wiring a few hundred dollars to a person they have never met, convinced they are helping a romantic partner through a crisis. They will do it again next week, and again the week after that. On average, they will send $2,020 before their bank notices anything unusual.
That figure comes from the Federal Trade Commission’s May 2026 consumer alert on imposter-scam trends, which found that reported romance-scam losses climbed 22 percent in 2025 compared with the prior year. Separately, the FTC documented $12.5 billion in total reported fraud losses for calendar year 2024, a figure that covers all fraud categories, not just romance schemes. Within that broader landscape, romance scams stood out as one of the fastest-growing subcategories of imposter fraud. And because most victims never file a complaint, the real toll is almost certainly higher.
The playbook has not changed, but it works better than ever
Fraud researchers call it the “grooming arc.” A scammer builds a profile on a dating app, social media platform, or messaging service, then starts a conversation with a target. The early phase is all emotional investment: frequent messages, personal confessions, rapid declarations of affection. The scammer might claim to be a military officer stationed overseas, a surgeon on a humanitarian mission, or an entrepreneur stuck in a foreign business dispute. The specifics vary. The function is always the same: explain why they cannot meet face to face or join a video call.
Once the target feels emotionally committed, the requests for money begin. They are framed as urgent and temporary. According to FTC complaint data, the most common payment methods are wire transfers, gift cards, and cryptocurrency, all of which are difficult or impossible to claw back once sent.
Critically, the money leaves in small, repeated amounts rather than one large transfer. A $200 wire to a new recipient does not look suspicious on its own. Neither does the $300 that follows, or the $500 after that. Each transaction slips under the thresholds that trigger automated fraud alerts at most banks. By the time the cumulative total crosses $2,020, the victim may have made a dozen transfers across several weeks, every one of them individually unremarkable to the institution processing it.
That drip pattern is the core of the problem. It is engineered to exploit the gap between what a victim experiences (a deepening relationship) and what a bank sees (a series of modest outgoing payments).
The FTC and FBI are both flagging growth, though from very different vantage points
The FTC is not the only agency watching the numbers climb. The FBI’s San Francisco field office issued a Valentine’s Day warning citing year-over-year growth in Internet Crime Complaint Center filings for its region. That alert also flagged evolving tactics, including the possible use of AI-generated text, synthetic voice, and deepfake video to make scammer personas harder to detect.
The two sources differ sharply in scope. The FTC’s consumer alert draws on a national complaint database covering all 50 states. The FBI release is a single field office’s seasonal press statement focused on the San Francisco division’s territory. They do draw on separate reporting systems: the FTC collects complaints through its own consumer portal, while the FBI’s IC3 uses a different intake process. The fact that both point toward growth is suggestive, but the FBI release is far too narrow to serve as independent national confirmation of the FTC’s figures.
Historical data adds context. A 2021 FTC Data Spotlight found that romance scams already produced the highest median losses among all fraud categories in 2020. The FTC’s own subsequent annual reports have shown continued growth in overall fraud losses since then, and the 2025 romance-scam figures are consistent with that broader upward trend, though individual year-to-year movements in the romance subcategory have not been independently verified for every intervening year.
What the numbers still leave out
For all the alarm in the headline figures, significant gaps remain. Neither the FTC’s 2026 alert nor the FBI’s warning breaks down the 2025 romance-scam cohort by victim age, gender, or the platform where contact began. Prior FTC reports have shown that adults over 60 tend to report the highest individual losses while younger adults file more total complaints, but those patterns have not been confirmed for the latest data.
The absence of platform-level detail is a particular frustration for consumer advocates. Without knowing whether the bulk of new cases start on dating apps, social media feeds, or encrypted messaging services, targeted interventions are largely guesswork.
The FBI’s reference to AI-assisted scams also lacks hard numbers. The San Francisco office warned that scammers may be using generative AI tools, but it did not attach complaint counts or loss figures to AI-specific cases. That distinction matters: policymakers weighing responses like mandatory watermarking on synthetic media need evidence that AI tools are changing outcomes, not just that the tools exist.
One methodological note worth flagging: the FTC’s 2021 Data Spotlight acknowledged that including or excluding certain transaction-level reporters changes the median loss calculation. The agency has not published an equivalent disclosure for its 2025 figures, so year-over-year comparisons should be read with that caveat.
If you have already sent money, speed is everything
For anyone who has wired funds to a person met only online, the recovery window is measured in hours, not days. Contact your bank or payment provider immediately. Then file separate reports with the FTC’s fraud-reporting portal and the FBI’s Internet Crime Complaint Center. Each feeds a different investigative pipeline, and neither automatically shares individual complaints with the other.
Save everything: screenshots of conversations, transaction confirmations, profile photos, phone numbers, and any identifying details tied to the scammer’s accounts. Even when money cannot be recovered, detailed reports help investigators connect individual cases to larger networks and refine the risk models banks use to catch similar schemes earlier.
For people who have not sent money but feel pressured, a few red flags are nearly universal. The person refuses to video chat or meet in person. They claim to be overseas for work or military service. Financial requests are framed as emergencies. And the relationship accelerates to declarations of love far faster than normal. Any one of these signals warrants caution. Two or more together should prompt a hard stop and a conversation with someone you trust offline.
Banks control the chokepoint, and they are barely using it
Financial institutions sit at the exact point where intervention could matter most. Banks and payment apps see outgoing transfers in real time, and many already run fraud-detection algorithms for other scam types. Applying those tools more aggressively to repeated transfers to new recipients, particularly when the pattern matches the romance-scam drip of small, escalating amounts, could shrink the window between a victim’s first payment and an institutional flag.
The United Kingdom has moved further on this front. Since October 2024, UK payment firms operating on the Faster Payments system are required to reimburse authorized push-payment fraud victims up to £85,000, a policy that gives banks a direct financial incentive to detect scams before money leaves. No equivalent mandate exists in the United States, where victims of authorized transfers have little legal recourse once funds clear.
More aggressive detection would mean more friction for some legitimate transfers. A customer sending money to a long-distance partner they have actually met might face an extra verification step. But the federal data suggest the current balance is wrong: too much money is leaving too easily, and the people losing it are among the least equipped to absorb the hit.
The 22 percent rise in romance-scam losses is not an abstract data point. It represents thousands of people who believed they were helping a partner, funding a shared future, or rescuing someone from danger, only to discover the relationship was engineered from the first message. Closing the gap between the first dollar sent and the first alert raised will take more than seasonal warnings. It will take banks, platforms, and regulators treating the pattern of small, repeated transfers to strangers as the high-risk behavior the data already shows it to be.
