Abstract Thursday Morning Keynote: Nicklas Lundblad

Lies, damn lies and privacy

Nicklas Lundblad

Deputy CEO of the Stockholm Chamber of Commerce

Abstract

The relationship between privacy and human computer interaction is fairly well understood. That there is such a relation is hopefully no longer in question, and there are a number of different research projects investigating the nature of the relationship in-depth. One basic assumption in some of these efforts seems to be that we can derive a number of design criteria from looking at privacy requirements and perhaps even legislation, that we then need to respect these and limit the design space that is open to systems designers in order to preserve or protect privacy.

The legal set of requirements then basically works as a framework within which design solutions need to be chosen. I would submit that this is a thoroughly erroneous way to think about the relationship between privacy and HCI, and that it assumes that one part in this deeply dynamic relationship is static. The idea of privacy is changing rapidly as an effect of design choices already made, and not even law with its considerable inertia is immune to the constant re-negotiation that is permeating the understanding of privacy in our society. Privacy no longer defines the set of possible design solutions. In fact, with the rise of social computing the relationship has almost become reversed. As computer systems morph into identity producing systems maintaining and supporting both identity formation and compartmentalization, the design of these systems shape the different forms of privacy available to us. The question should no longer be how we design privacy into our computer systems, but rather how we design identity with our computer systems.

The term identity management is on several levels a misnomer. Social technologies do not manage identities but produce and enable them. Much as producers and consumers are turning into Benklerian prosumers we see identity production and privacy protection turn into what we for lack of a better term will call (ugh!) identacy prodection, the mutual production and protection of a continuous spectrum of identities in different contexts.

This new practice is the simultaneous production of identity and protection of privacy: by divulging information on the internet users are surrounding themselves with clusters of probable but mutually incoherent propositions about them. This new Nietzschean privacy paradigm where we "speak much to hide ourselves" offers what I believe could be interesting new challenges for HCI-research.

I will attempt to argue that some of these challenges have to do with enabling "social lying". By building lie-amplifying technologies and user interfaces and systems that support and strengthen lies we can support this strategy and turn from a tinkering practice into a systems feature. Lie-amplification technologies necessitate looking closely at the social practice of lying and how it is organized. The crucial thing is to enable lying that is not trust-erosive, but privacy conducive. This is no mean task and will require lots of thoughts about how lies can be used to provide privacy effects that can go beyond what we can produce today.

In addition to this I think we need to understand trivial anonymity and how this form of context can be revived on the internet. Trivial anonymity is the kind of anonymity we enjoy in situations where no-one knows us or cares about us: in a shop, on the train or on a flight. The thin identity-like veneers we need in these situations provide even better anonymity than if we were, say, flying in masks. One reason behind this is that we tend to forget these encounters quickly, and that they leave us with little or no impression that we can use. If we could replicate this trivial anonymity on the web, ensuring that users do not need to signal that they are making conscious efforts to be anonymous, much would be gained. Otherwise the signalling effects of using, say encryption, enable traffic analysis and have privacy eroding effects on their own.

Short Bio

Nicklas Lundblad is deputy CEO of the Stockholm Chamber of Commerce. He has a PhD in informatics from Gothenburg University, and is a member of the Swedish Government's ICT-council. He has written three popular books on ICT and public policy and was elected ICT-person of the year 2008. He most recently comes from Google where he served as European Policy Manager for two years, working with public policy in the Nordics, Baltics and Google's work with the OECD, ICC and IGF. Prior to this he co-founded a current affairs magazine in Sweden and started the Chamber of e-Commerce.

Dr Lundblad is an Eisenhower Fellow (2006) and a member of the board of the Swedish Research Institute for Legal Informatics at Stockholm University, where he got his law degree. He really likes writing his own bio, since he thinks it is neat to be able to get away with all different kinds of things, and sincerely believes that anyone who is talking about lies should be distrusted. Seriously, folks!