Snapchat Wants To See Everything Your Camera Sees

The most common lie of the 21st century is surely “I have read this agreement and agree to the terms and conditions.” To gain access to apps and social networks, we often breeze over the formalities to get at whatever ‘free’ service is purportedly being offered, signing over permission to view and analyse our habits, friend lists and interests. For much of the information, we’re happy to give it as we see it as inconsequential and worthless, whether because we don’t understand its importance, or its value isn’t obvious.

Those who warn about the dangers of oversharing, who tape over their computers’ webcams for fear of surveillance – are they really the tin hatters of our generation, or its watchmen?

A patent filed earlier this month by Snapchat reveals that, in order to give you more ads you’re likely to click on, the company is aiming to gain access to your smartphone’s camera and pictures.


Patent # US20160203586

Their patent application describes a photo filter engine which will scan your snaps for recognisable imagery, then offer you related filters to use in the app. Specifically, the engine will make “associations between an object and a source of image data, for example, a brand of a merchant in which case the associated photo filter may include images associated with the brand of the merchant.”

For example, a proposed use will be to overlay an image of King Kong if users point their camera at the Empire State Building. Reportedly, merchants, restaurants and individual users will be able to submit image filters for recognised objects.

Early uses of a similar technology in the app have been programmed to recognise certain objects, for example taking a picture of a coffee cup would result in you being offered photo filters featuring advertisement from coffee brands.

One interesting application mentioned in Snapchat’s patent is similar to, and yet arguably even more impressive than, Google’s camera translation capabilities: “A photograph including an object recognised as a restaurant may result in the user being presented with photo filters that overly a menu of the restaurant on the photograph. Or a photograph including an object recognized as a food type may result in the user being presented with photo filters that let the user view information e.g., calories, fat content, cost or other information associated with the food type.” Years from now, we may wonder how we ever did without it.

Targeted advertising in this day and age is becoming increasingly intrusive, and machine learning advancements are going to make it even more so. Snapchat has been struggling to monetise its rapidly scaling platform for a number of years, and this new technology could provide an effective way to do that. It’s certainly not ideal, but it’s far less intrusive than mobile popup ads.

Of course, you needn’t move to the woods and don your tinfoil just yet, but it might be worth paying attention to which of your devices have access – and right to view – to your microphone, contacts, camera and search history, for a start.