Features

net.wars: Applesauce

Modern life is full of so many moments when you see an apparently perfectly normal person doing something that not so long ago was the clear sign of a crazy person. They're walking down the street talking to themselves? They're *on the phone*. They think the inanimate objects in their lives are spying on them? They may be *right*.

Last week's net.wars ("The open zone") talked about the difficulty of finding the balance between usability, on the one hand, and giving users choice, flexibility, and control, on the other. And then, as if to prove this point, along comes Apple and the news that the iPhone has been storing users' location data, perhaps permanently.

The story emerged this week when two researchers presenting at O'Reilly's Where 2.0 conference presented an open-source utility they'd written to allow users to get a look at the data the iPhone was saving.

But it really begins last year, when Alex Levinson discovered the stored location data as part of his research on Apple forensics. Based on his months of studying the matter, Levinson contends that it's incorrect to say that Apple is gathering this data: rather, the device is gathering the data, storing it, and backing it up when you sync your phone. Of course, if you sync your phone to Apple's servers, then the data is transferred to your account – and it is also migrated when you purchase a new iPhone or iPad.

So the news is not quite as bad as it first sounded: your device is spying on you, but it's not telling anybody. However: the data is held in unencrypted form and appears never to expire, and this raises a whole new set of risks about the devices that no one had really focused on until now.

A few minutes after the story broke, someone posted on Twitter that they wondered how many lawyers handling divorce cases were suddenly drafting subpoenas for copies of this file from their soon-to-be-exes' iPhones. Good question (although I'd have phrased it instead as how many script ideas the wonderful, tech-savvy writers of The Good Wife are pitching involving forensically recovered location data). That is definitely one sort of risk; another, ZDNet's Adrian Kingsley-Hughes points out is that the geolocation may be wildly inaccurate, creating a false picture that may still be very difficult to explain, either to a spouse or to law enforcement, who, as Declan McCullagh writes know about and are increasingly interested in accessing this data.

There are a bunch of other obvious privacy things to say about this, and Privacy International has helpfully said them in an open letter to Steve Jobs

"Companies need openness and procedures," PI's executive director, Simon Davies, said yesterday, comparing Apple's position today to Google's a couple of months before the WiFi data-sniffing scandal.

The reason, I suspect, that so many iPhone users feel so shocked and betrayed is that Apple's attention to the details of glossy industrial design and easy-to-understand user interfaces leads consumers to cuddle up to Apple in a way they don't to Microsoft or Google. I doubt Google will get nearly as much anger directed at it for the news that Android phones also collect location data (the Android saves only the last 50 mobile masts and 200 WiFi networks). In either event, the key is transparency: when you post information on Twitter or Facebook about your location or turn on geo-tagging you know you're doing it. In this case, the choice is not clear enough for users to understand what they've agreed to.

The question is: how best can consumers be enabled to make informed decisions? Apple's current method - putting a note saying "Beware of the leopard" at the end of a 15,200-word set of terms and conditions (which are in any case drafted by the company's lawyer to protect the company, not to serve consumers) that users agree to when they sign up for iTunes – is clearly inadequate. It's been shown over and over again that consumers hate reading privacy policies, and you have only to look at Facebook's fumbling attempts to embed these choices in a comprehensible interface to realize that the task is genuinely difficult. This is especially true because, unlike the issue of user-unfriendly systems in the early 1990s, it's not particularly in any of these companies' interests to solve this intransigent and therefore expensive problem. Make it easy for consumers to opt out and they will, hardly an appetizing proposition for companies supported in whole or in part by advertising.

The answer to the question, therefore, is going to involve a number of prongs: user interface design, regulation, contract law, and industry standards, both technical and practical. The key notion, however, is that it should be feasible – even easy – for consumers to tell what information gathering they're consenting to. The most transparent way of handling that is to make opting out the default, so that consumers must take a positive action to turn these things on.

You can say – as many have – that this particular scandal is overblown. But we're going to keep seeing dust-ups like this until industry practice changes to reflect our expectations. Apple, so sensitive to the details of industrial design that will compel people to yearn to buy its products, will have to develop equal sensitivity for privacy by design.