November 18, 2026

The Face on the Camera Isn’t Always the One in Front of It: How Fraud Crews Are Beating Online Identity Checks

Entrust’s newest identity fraud report finds deepfake detections up 165% in a year, with criminals using virtual cameras and device emulators to slip fabricated or stolen faces past selfie and ID checks. Here is how the scheme works, why it has become an organized trade, and what it means for businesses that verify people online.

A young woman at a kitchen table looking closely at a smartphone beside an open laptop

Picture a new customer signing up for a business account. The app asks for a photo of a driver’s license and a quick selfie video, the face matches the document, and the account opens. In a growing number of cases, according to Entrust, the person holding that phone never existed, or did exist but never agreed to any of it. The company’s new report, The Professionalization of Fraud: 2027 Identity Fraud Report, describes how identity checks have become a production line for criminals. The summary below is based on Help Net Security’s coverage of the report.

Making the face is only step one

Entrust says deepfake detections rose 165% year over year, as criminals use AI-generated IDs, photos and videos to pose as real customers. Those attacks can be personalized, automated and repeated against many services at once. But a convincing fake is worthless until it gets through the verification process, and that is where the newer tricks come in.

In an injection attack, the fraudster never points a real camera at anything. A virtual camera feeds fabricated footage into the verification software as if it were a live recording, and device emulation software imitates the phone or computer submitting it. Stolen photos and videos travel the same route. As the report notes, an image can be completely genuine and still be submitted by someone with no right to use it. Entrust’s measurements through device intelligence, which examines the camera, device and session behind each submission, show the fraud rate rising 42% year over year.

Paper fakes have not gone away

Document fraud is still the biggest category of attack in Entrust’s data, even though its rate fell 19% from the prior year. Physical counterfeits made up 61.1% of fraudulent document submissions, and medium-sophistication alterations and counterfeits, the kind that need a closer look to catch, accounted for nearly 39%. Simon Horswell, a senior fraud specialist manager at Entrust, says criminals are investing in better templates, stolen credentials and specialized skills, with AI helping them scale up.

When the real person is the weak link

Not every attack needs a fake. A customer coached by a scammer can pass verification with their own genuine ID and face. An employee can be talked into approving a transaction or an action that opens the door to fraud. Entrust describes criminal groups dividing the work, with different specialists handling different stages and coordinating people, tools and data so the same playbook can be repeated across a customer relationship.

Where it is hitting

Financial services draw the most attempts, because a verified account offers access to money and can be reused for money laundering, mule activity and further fraud. By region, the Americas had the highest document fraud rate and the fastest growth in deepfake attacks, with driver’s licenses the most targeted credential, including those from Texas, California and Florida. Europe had the largest share of deepfake attacks in Entrust’s dataset, with detections up 462%, and Asia-Pacific saw the highest proportion of digital forgeries.

What this means for your company

You may not run a bank, but the same tricks reach smaller firms in three ways. If you onboard customers, tenants, drivers or contractors online, a single selfie-and-ID check is no longer proof on its own; ask your provider whether it detects injected video and emulated devices, and add a second signal for higher-risk accounts. Hiring is another exposure, since remote interviews and onboarding can be faked the same way, so verify identity again at the point of granting system access. And your own staff remain a target: build a rule that payment changes, password resets and new account approvals are confirmed through a known phone number or a second person, not the channel the request arrived on.

If you are unsure how well your onboarding, hiring or help desk checks would stand up to a convincing fake, we can help you test them and add the right safeguards. Reach out to Layer 7 Data Solutions to discuss how we’ve helped companies similar to yours.

About The Author