The Ins and Outs of Opting
There has been another rash of security/privacy issues by major internet companies. Actually it’s more of an ongoing issue than it is a recent outbreak. And much of the ongoing trouble is related to a poor understanding of “opt in” vs “opt out” methodologies, and some pretty poor business choices in that area.
Google just announced that wireless network owners can now “opt out” from its Wi-Fi geolocation map database. Many have greeted this as good news and responsible behavior on Google’s behalf. Others, myself included, view this as a classic case of a business doing essentially nothing to change it’s behavior, and then promoting the non-effort as a valuable security benefit to their customers and the world at large. Google believes that once you’re using any of their services, you’ve essentially opted in to anything they want to do. More on that in a minute.
Another consumer favorite, Facebook appears to be tracking 90 days of everything it’s users (and some suggest even former users) browse on the web. This is beyond just tracking what you’re doing inside Facebook itself. And there are also allegations over whether or not they actually are storing profile information about people who have not even joined Facebook. This is another company that believes in a policy of opting you in to anything they want and then letting you opt back out. They know that a lot of people aren’t savvy enough to understand, others too lazy, and others will never even be aware of the issues.
Verizon tracks everything you do with your phone, so do pretty much all the cell phone companies. Recently Verizon started allowing people to opt out. Josh Constine at TechCrunch mentions that at least they don’t call it “Greater Choice” like Google does. But his take is everyone is saying “Why can’t we be evil too?”
Strangely enough, AT&T takes the opposite move of letting people opt in. Pretty ironic for a company who’s logo resemble the Death Star, but commendable.
The problem with “opt out” is that it works well outside of privacy areas. It also works in areas where you have an explicit relationship. For example, if I create a Google account it will keep track of what I search, unless I opt out. Most companies that have web accounts work in this way, for example with their email lists. This is a very reasonable method – you contacted me, so you don’t mind if I contact you. You see this normally as little “send me your junk email” boxes. You can judge the company based on whether the boxes are clicked or empty by default on their sign-up forms.
The stakes for things like this are low – the worst case is that some web site sends me a bunch of junk email, and if they’re a responsible company, they’ll respond to my “stop that” request.
The difference with privacy issues is that the stakes are much higher, and the awareness is much lower. If someone decides that by using their website I agree to let them track my every move on the web, it’s unlikely that I’ll ever figure it out. And they may end up being privy to something I didn’t want to share with them. Opting people in by default to such things is unethical behavior at best. What’s the rational connection between me using your website and me giving you permission to spy on all my web activities? There is none of course.
In the case of the Google Wi-Fi mapping they’re collecting your data whether or not you have a relationship with them. This is one step worse than the Facebook issue. In this case they’re literally driving the streets of the world looking for Wi-Fi (we used to call this warchalking) and then adding you to a database.
You may not even be aware they’re collecting your data. In fact, the odds that homeowners ARE aware are extremely small. And yet they’re using on opt out methodology, just to cover their butts. Which essentially means that they’re opting you in to something, without your permission, without your awareness. And they justify that because their company motto is “Do No Evil”.
The truth is that it’s a very questionable practice to collect someone’s information without their knowledge. If they want to build a database, then can simply switch to an opt in method. Instead of my changing my SSID if I happen to know that they might drive by someday, (which is inconvenient because I have to reset all the devices using my network, including frequent guests devices) they can go to a method where they only collect data from those who indicate willingness by changing the name.
Instead of changing my SSID from “mynetwork” to “mynetwork_nomap” to opt out, I should be able to change to “mynetwork_map” to opt in. Anyone who doesn’t want it doesn’t have to do anything. Anyone who is unaware will not be unintentionally opted in. Anything less is not only unfriendly to consumers, it’s just plain evil.
- » AWS will support NVIDIA’s T4 GPUs focusing on intensive machine learning workloads
- » Addressing cloud sprawl: Combining security best practices with business foundations
- » Enterprise demand for agile, data-centric architectures: The next wave of big data and analytics
- » Practical cloud considerations: Security and the decryption conundrum
- » Joining the dots: Equipping the hybrid cloud to meet the needs of the digital enterprise