AI Deepfakes Drive Record Romance Scam Losses on U.S. Dating Apps

Romance scam losses in the United States have reached record levels, driven by a surge of AI-generated profile photos that fool even experienced online daters. According to the Federal Trade Commission, Americans lost $1.14 billion to romance scams in 2023, a figure fraud researchers say has continued to climb through 2025 and into 2026 as generative AI tools have become freely available.

Unlike stolen photos — which can be caught by a simple reverse image search — AI-generated faces are unique images that have never appeared anywhere on the internet. That makes traditional detection methods largely ineffective, and has pushed the burden of verification onto new tools like reverse face search.

Industry survey data published in early 2026 found that 61% of active online daters reported being deceived by a fake profile, and 84% said AI deepfakes had made it harder to trust new matches. The average reported loss per victim reached $10,000, with a significant share of victims losing over $50,000 before realising the relationship was fabricated.

The FTC's most recent Consumer Sentinel data places romance fraud in the top five fraud categories by total dollar losses, ahead of online shopping scams and impersonator fraud in many states.

Before AI image generation became mainstream, scammers typically stole profile photos from real people's social media accounts. That method left a traceable trail: a reverse image search could return the original source and reveal the stolen identity. AI-generated faces close that gap entirely.

Tools including Midjourney and Stable Diffusion can produce a photorealistic face that is unique, has never been indexed, and passes visual inspection by most users. The only reliable check is reverse face search — not reverse image search — which looks for the same facial geometry across the public web rather than matching pixel patterns.

Major dating platforms have begun rolling out identity verification features in response. Tinder announced the expansion of its Face Check liveness verification to all U.S. users in late 2025, requiring a video selfie that is compared to the profile photo in real time. Bumble and Hinge have deployed similar features, though participation remains optional on most platforms.

Consumer advocates argue that voluntary verification is insufficient, and have called on the FTC to mandate baseline identity checks for dating applications with more than one million active users.

The FBI recommends running a reverse image search on any profile photo that raises suspicion. For AI-generated photos that return no image search results, a reverse face search — which checks whether the same face appears elsewhere on the public web — is the more effective next step. A face with no public footprint at all is a meaningful red flag when combined with other warning signs such as avoidance of video calls and requests for money.

Any contact that introduces investment links, crypto platforms, or payment apps should be reported to the FTC at reportfraud.ftc.gov and to the platform directly.