The Insurance Company Will Webcam You Now

Image recognition firm Facebanx is introducing webcam identity verification for online insurance sites, banks, commerce, and gambling.

Odds are that your computer has a webcam. Even if your computer doesn’t have a camera, your mobile phone definitely does. But that camera is good for more than just Snapchats and Skyping. Online commerce giants, banks, financial firms, and other sites are keenly interested in using webcams to reduce fraud–they believe that if they can visually identify their customers, credit card fraud will be reduced. One British company, Facebanx, believes they can make this part of daily Internet life.


Facebanx’s CEO, Matthew Silverstone, is offering websites and others backend software that compares a livestream of a person’s face to an image submitted to the company. Companies using Facebanx’s software are required to alert users their webcams are being activated, and users voluntarily submit their images to the companies. In the United Kingdom, Facebanx is being used for pilot projects which will reduce insurance premiums in exchange for account holders submitting to image verification as shown in the video below.

In an interview with Fast Company, Silverstone emphasized the fact that webcam or mobile photos are a simple way for sites to add an extra level of verification to user logins. “We are another weapon in the fight against fraud–no fraudster wants to be identified visually,” Silverstone said. Potential corporate customers are also interested in the deterrent factor of Facebanx’s technology; by requiring (or at least encouraging) user pictures on record, they hope to reduce the amount of fraudulent transactions or account hacks taking place.

When users’ pictures are uploaded to Facebanx’s servers, a still picture of the user’s face is captured from the camera’s live stream or from mobile camera stills. Algorithms are applied to the images to verify they are indeed legitimate photos and that they were not previously taken. Facebanx also compares the photos against an updated database of photographs to flag possible fraud.

Once an image is flagged as suspicious, Facebanx’s backend software can also initiate real-time intervention from a client’s consumer service or anti-fraud representatives.

The company is a spinoff of OhHi, a video chat startup specializing in event streaming and teleconference services. According to Silverstone, the company first began developing the facial verification concept in 2012. Facebanx launched their services in April of 2013.

Although primarily aimed at the insurance industry, one early client integrating Facebanx’s verification technology is a European stock trading site. Other clients who show interest include gaming/casino sites, online banking sites, and e-commerce sites. The backend software Facebanx markets requires no downloads by users and can be used on regular PCs, tablets, and phones.


Video verification is big business for banks, insurance companies, commerce sites, and online gambling. Adoption of video verification has been slow for several reasons. Facial recognition algorithms–especially cross-platform ones which can be applied to Macs, Windows/Unix machines, and mobile devices–are still buggy and take precious seconds to verify. Although large corporations Fast Company spoke with are keenly interested in online facial recognition, a combination of privacy concerns and worries about functionality mean that the tech is still largely in testing mode.

Nonetheless, the Federal Bureau of Investigation is investing a cool $1 billion into developing facial recognition technologies for law enforcement. HSBC is also testing facial recognition in a British test project, and iPad POS developer Revel is introducing a Facebook photo verification pilot for retailers. Due to differing privacy laws in Britain and the United States, facial recognition tech will be harder to implement stateside than in the United Kingdom.

In the meantime, privacy activists are already working on ways to foil facial recognition software.

[Top Image: Steve Jurvetson]