Robin Wilton

If you have five minutes, it’s well worth spending them on a quick read of this blog post by Timothy O’Neil-Dunne. Don’t let yourself be put off by a couple of spelling glitches, because the post raises a number of key themes for online privacy.

For instance, he gives a couple of good examples of how Privacy Policy Statements can actually have nothing to do with protecting your privacy as a user of the service, and everything to do with covering the service provider’s backside no matter what they choose to do with your data. There will be more on that topic, incidentally, when my first Gartner research report comes out… so watch this space.

O’Neil-Dunne also points out how much the onus is on you, the end user, to manage your own privacy online – not unreasonable in principle… after all, it’s always good to encourage people to take responsibility for their own safety – but in practice, as he says, the tools and mechanisms at our disposal are few and weak.

He goes on to give the example of Rapleaf’s opt-out option. Rapleaf have been in the spotlight this week as one of the third-party entities who consume much of the data about you which, knowingly or otherwise, you disclose via sites like MySpace, Facebook and others.

In principle, as I say, there’s nothing wrong with expecting people to take responsibility for their own safety when navigating the web, just as we expect people to take responsibility for their own safety when crossing the road. However, as that analogy would suggest, it’s also wrong to suggest that people are the only ones with any safety responsibilities online. Think of all the other factors which have a bearing on your safety when you’re crossing the road: other road-users have strict codes of conduct to adhere to, and in the case of motor vehicles, they have to display a unique identifier prominently to public view; there are architectural safety mechanisms such as road layout, road markings and signage; and there are operational mechanisms such as traffic signals and pedestrian crossings. All these are backed up by legislation and a substantial investment in enforcement, with case law and penalties laid down for when something goes wrong.

Most of those statements do not have an online privacy analogy.

The opt-out mechanism for RapLeaf is a good example of how the principle of user responsibility is undermind by poor tools – and I’m not singling RapLeaf out in this respect: this is a general point.

1. If you don’t want RapLeaf to aggregate data about your behaviour, you have to agree to store a cookie saying so. For that to be a realistic option, of course, you have to be aware of RapLeaf’s existence in the first place – and like many members of the targeted advertising ecosystem, their existence is generally entirely invisible to the individuals whose personal data they consume. It’s hard to guess that, for your own good, you ought to be opting out of something you don’t know is being done to you, by an organisation you’ve never heard of.
2. Opt-out based on a cookie requires you to enable cookies in the first place. If you’re particularly privacy-conscious, you probably either disable cookies in your browser, or instruct your browser to delete cookies each time you close it. Service providers dislike the first of those options so much that, if you try it (or if you ask to be notified each time a cookie is set) you’ll soon conclude that the Internet is unusable under those conditions. If you go for the second option, then of course every time you re-start your browser, you’ll have to remember to visit RapLeaf before going anywhere else, and re-set your opt-out. Oh, and do it for any other organisation you want to prevent from doing similar things… whether you’ve heard of them or not. Completely impractical.
3. What I’ve described above is pretty much the ‘best case’ scenario; it assumes that you’re using a PC-style browser, with quite good controls for letting you specify how it should handle cookies. But – and be honest here – how much cookie management have you done on, say, your iPod/iPhone or other palm-top web client? Do you even know where the cookie management functions for Safari are on an iPod? Even if you’ve found them (and if you haven’t, here are some pointers), you will have noticed that they are far less granular and rich in function than the browser equivalent. In fact, they are pretty much binary. You can accept cookies or not; you don’t get to specify which sites you are happy to accept them from and which you would like to block. You can keep all your cookies or delete them all. You don’t get to specify (or even see) individual cookies to keep or discard.

So, when you read point (2): you may have concluded that the best browser strategy would be to disable cookies for all sites except those which require you to accept an opt-out cookie. 10/10 for logic, 0/10 for portability to devices like iPods. That strategy just isn’t an option on that platform.

It has come to something, hasn’t it, when the only available mechanism for saying that you don’t want to be tracked through cookies is… a cookie. It essentially makes a mockery of the idea that users can take control of their own privacy online, because it makes opting out entirely unworkable.

I’d be interested to know whether you think there’s a solution to this. There are those who argue that banning cookies would kill the web – both in terms of interactivity/usability, and in terms of commercialisation. Similarly, there are people who maintain that insisting of opt-in as a default rather than opt-out would kill many of the most innovative business models out there.

Is there a compromise or sweet spot which avoids those pitfalls while still offering users meaningful ways to take responsibility for their own online destiny?

2 Comments »

Category: Uncategorized     Tags:

Share

Tweet