Your mobile phone leaks

Published in design/architecture magazine ICON, issue 106 on mobile phones:

This year the number of mobile phones will exceed the 7 billion humans on the planet. For this issue we asked novelists, academics, experts and designers to reflect on this communication revolution, in a 22-page special on how cell phones have changed the ways we behave, connect to and navigate the world. And to make their own predictions about how mobile phone technology will look in the future …

 

Your mobile phone leaks. Behind the user interface, out of immediate view, it’s sharing a lot more data than many people realise.

Take location. In exchange for offering Google Maps as a free service, Google extracts the price of knowing where your phone is at all times, even when the app isn’t running. Your home and work addresses are easy to identify (your habitual locations at 3am and 10am respectively). These can be cross-referenced against MOSAIC (market research company Experian’s consumer classification) or Zoopla house price records to transform location into income and demographic data, allowing users to be sold as micro-targeted ‘market segments’ of high value to advertisers.

Mobile web surfing habits provide another stream of data. Mobile operators use deep packet inspection and redirect mobile web traffic through their own servers to manage network performance, but this also allows them to monitor the websites people visit. Private internet use through VPNs may also be constrained, allowing fewer channels for private browsing – and child protection agreements mean that everyone not verifying their identity as over-18 will be blocked from much of the web. Legally operators must enforce blocks on a small blacklist of domains (e.g. child pornography), but monitoring web history is also data that is highly commercially exploitable.

Information storage is increasingly cheap and data protection laws some distance behind the technology, meaning that companies are building the biggest possible datasets now to hedge against future restrictions.

Less legitimately, mobile phones can also easily be compromised by malware and spyware. Apps may ask for greater rights than they strictly need, allowing remote access to the phone’s microphone and camera, and sharing text entered (e.g. emails, passwords) and location data. Occupy London protestors have been known to remove batteries and keep mobiles in a separate room while meeting to plan future actions. This may seem paranoid, but the Mark Kennedy case has shown police infiltration of ‘domestic extremist’ groups to be commonplace.

Does mobile data sharing matter? Some would argue no: users are knowingly exchanging their data for free access to entertaining and useful services. But the impact of such bargains goes beyond the individual. Companies such as insurers and financial lenders are keen to use whatever data they can to minimise risk. This may mean denying insurance or a mortgage on factors outside the applicant’s control – simply the likelihood that “people like you” (by location, or web use) are more likely to default on payments.

The customised advertising enabled by mobile data also have their costs. By being delivered on the basis of aggregated and probabilistic data, the recommendations made are normative. Does the working class teenager see ads for jobs in McDonalds rather than university degrees? Is pregnancy advice limited by religious affiliation? Personalised services offer convenience at the price of potentially constraining our possibilities for action.

Behind the commercial value of mobile data is network analysis: modelling our social relationships (call histories, social media friends) as the nodes and links of a graph, and analysing patterns and clusters. This has substantial predictive capacities: where one user is unknown to a mobile operator (or to Facebook), many personal details can be inferred from their patterns of interaction with known entities. An individual does not have to be directly known to be present in the network through their relationships with others.

Social media analysts do not only focus on the ‘social graph’ of relationships between people – they analyse ‘interest graphs’ (relationships between profile interests or topics of discussion, e.g. music or technology) in exactly the same way. To what extent does “the individual” remain the primary unit within these assemblages of behavioural data, social, material and semiotic relationships?