RMM

(Last week Tom asked me, as a part of his current series of blog posts, to consider the ethical ramifications of making scheduling information available through social media. What follows below is an attempt to begin to think more generally about the ethics of social media, specifically the essential notion of privacy.)

From Facebook’s almost surreally invasive ‘Beacon’ project to more general concerns about just how much Google & co. know about our digital habits, there has been much concern of late about the idea of privacy in online social media. Indeed, there has even been an inquiry into it in the US House of Representatives (back when they didn’t have $700 billion of corporate welfare to distribute). Yet rarely has there been any focused discussion of the underlying ethical questions surrounding online social media - specifically what is at stake when Google et al. are able to track my online activities and sell that information to advertisers.

David Rostan, the director of Listens to You (a new company which lets users chose what ads are targeted at them - and a company I’ll be profiling in more detail this month) put the whole issue of targeted ads into context for me in an email:

Why won’t people let AT&T (the US phone company) tap their phones and serve ads based on what they hear, but they allow essentially the same thing online?”

Indeed, we won’t even let the police listen in to our calls without a warrant, so how come Google can listen to my g-chat? The companies who deal in targeted advertising would argue that their listening in is ethical because it meets the three central criteria for any ethical interaction: informed consent, beneficence, and justice. (More after the jump…)

So:

a) the users of Facebook etc. agree to this when they agree to the terms and conditions. File under the ethical category of ‘informed consent’ (ie: people know what they are getting into)
b) the users are receiving a service (social media) for free, which must be paid for in some way, in this case by targetted advertising. File under the ethical category of ‘justice’ (ie: the fair distribution of risk and benefit)
c) the users are not harmed by this, indeed they benefit from products and services. File under the ethical category of ‘beneficence’ (ie: do good/do no harm)

Now, we can make the opposite case again using those same criteria:

a) most users do not read the terms and conditions properly. Indeed, a report form Carnegie Mellon University suggests that it would take a whole work week (42 hours) to read all the online privacy policies for every website we use in a year. We could argue that, as beneficiaries of the contract, the social media providers have a *duty* to ensure that users actually do read the T&C. In the absence of this assurance, we can argue that there is actually an absence of informed consent in the use of social media data.

b) If something is explicitly offered without charge, as social media are, then there is no obligation upon the user to ensure that the provider profits from the provision. Indeed, in profiting from the provision of a supposedly ‘free’ service, we could further argue that the companies have misled the users: more lack of informed consent. This also undermines the ‘justice’ criterion: there can be no discussion of the distribution of risk if something is given away for ‘free’.

c) it is possible to argue that users *are* harmed by targeted advertising. The obvious example is the invasion of privacy (see Facebook’s disastrous ‘beacon’ fiasco for the perfect instance of this). Privacy is invaded both personally (in your own interaction with the social media service) and publicly (in the way in which your information is shared with people you do not know).

There is a further, more subtle reduction of privacy made possible by targeted ads (and I’m grateful to my colleague here at Johns Hopkins, Dr Alan Rubel, for this idea) - the information provided by social media users is ‘profile’ information, about themselves. This is not necessarily the right information to use when deciding what sort of goods and services somebody wishes to be sold. Indeed we might argue that it is very much the *wrong* sort of data. In that case, the ‘profile’ that the social media service builds of you (and sells to advertisers) is, in some ways, false (we’ve all gone ‘wuh?’ when a targetted ad seems really out of place). Now, whilst that’s annoying (a violation of the beneficence criterion), Dr Rubel would also argue that the dissemination of false information is a reduction of privacy, as you have the right to maintain a public persona of your own choosing, not of the social media company’s design.

Let’s go one step further and talk about teleology: the aims of the provision of personal information. Now, if you have read the terms and conditions and agree, with fully informed consent, then there’s no ethical problem with the company sharing your information, right? Well, it depends on what you believed to be doing when giving that information. Sure, you agreed to having advertisements aimed at you. But does that necessarily mean that you agreed to having a marketable, saleable online profile of your identity created? The two are rather different things.

The key thing here, I think, are the terms and conditions. The companies’ argument - that there is informed consent because we tick that little box - is what I would call ’soft ethical cover’. There’s no active effort to ensure informed consent, merely the provision of the (unused) opportunity to inform yourself (as a user) of what you are consenting to. If the intent of the company is to use your data for profit and they actively inform you of this, then there’s no ethical issue. But I think that the general failure to actively ensure informed consent, rather than simply provide the opportunity for it, renders the use of the data unethical.

But it is legal. Simple contract law. You tick the box, then in law you read the terms and conditions, no question.

Ethics is not the law.

A final thought, bringing things back to Tom’s blogs about scheduling. The two types of data in question here may be ethically different. ‘pProfile data’ of the sort now collected is things like your age, what you buy, what you listen to, etc. The sort of ’scheduling’ data Tom is talking about is where people go, who they see, talk to etc. We could say that there is a category difference and that rather than simply collating data for sales purposes, the collection of scheduling data constitutes active surveillance rather than market research.

This is just some initial thoughts on the ethics of online social media with regards to privacy. Later, I’ll be profiling the company Listens to You, as a way of providing some concrete examples of how digital social media can operate ethically and profitably.