A colleague recently brought up an interesting discussion — how to hide from Google?
Privacy and anonymity are good things. Bruce Schneier has some great points – they are essential for democracy, and absence of anonymity is ripe for other kinds of abuse.
All arguments against online privacy and/or anonymity come down to the question, “what do you have to hide?” to which the only useful response is, “nothing, let me film you the next time you use the toilet”.
My first instinct is, “why bother”? As others have pointed out, staying anonymous online is a myth. Even if you disconnect yourself from Facebook, webmail accounts, and any support forums you might otherwise use, and religiously heed the dictum that you should never post anything anywhere online that you might not want someone to see, a sufficiently motivated analyst would find what I would refer to as second- and third-tier records, such as credit card bills, bank accounts, social security numbers, driver’s license details, etc. — more than enough for most purposes.
Both privacy and anonymity are, at most, best-effort affairs. So what is the motivation for this sort of research? A few plausible scenarios:
- Identify theft (e.g. constructive, for financial or other gain, or destructive, for defamation/slander)
- Targeted investigation (e.g. for a job, political office, or crime-related research, etc.)
- Statistics (e.g. targeted marketing)
- Trawling (e.g. to find instances of wrongdoing)
- Other (e.g. digging up dirt, fun)
I would guess that a vast majority of information collection falls into category 3, followed at a distance by category 4. Category 3 is worrisome mainly if you disagree with it out of principle. Fair enough. Growing technological capabilities of data cross-correlation systems allow for easy creation of entire profiles from otherwise disconnected bits of information. The potential for abuse of such a holistic, instantly accessible set of information about you as an individual is immense.
Even in the “right” hands, such as law enforcement (depending on your point of view) and responsible marketing types working under a strong privacy policy, the temptation to misuse data is inevitable. It is human nature to break rules, even in small ways — we are not automatons. I won’t go into scenarios about what even a government agency, scrupulously following all laws to the letter, could do with instantaneous, 24/7 knowledge of what you do, where you are, and what you’re thinking (category 4, followed by category 2), let alone a criminal element with stolen data about you (categories 1 and 2). Just in the hands of advertisers alone, the potential for irritation is limitless.
So let’s say that, out of principle, you object to the collection of information in the course of routing activities like online searches and use of free webmail service. I firmly believe that the best way for individuals to make the collection and correlation of information more difficult is to create as much entropy as is feasible. You can’t make it impossible, but you can at least make it not as easy.
In addition to at least making an effort to secure your Facebook profile as much as possible, some basic points:
Use GPG or PGP for mail and files. Most mail clients have easily available plugins. Make your key available.
When using free services like (highly recommended, by the way) Dropbox, use something like PGPdisk. You won’t be able to access it via the web interface, but it provides good security for files you share between computers.
Use SSL whenever you can. Even if you don’t use a commercial ($$) certificate for your own web page, you’re just trying to encrypt traffic. I recently made the mistake of accessing one of my (non-SSL) website accounts from a public wireless network, and only realized my mistake two seconds after logging in. Whoops. Password change time…
Use tor, and the excellent Vidalia package (on the tor site). Tools like the Firefox torbutton add-on make this a snap.
Tunnel things via SSH, and only use SSH for access to UNIX boxes. Of course, you’re already doing this.
Search on googlesharing (assuming you trust Moxie Marlinspike more than you trust Google) and use any of the other anonymizing services out there. If you run your own system, I can also recommend tools like nph-proxy.
And, of course, never post anything online you wouldn’t want someone else to see.
But again, full anonymity and privacy are illusory. Someone who cares enough can always find out about you, given sufficient tools and determination. Enter a colleague’s attitude, which I find pretty interesting.
His point is that it’s best to establish patterns of regular usage, that most of what you do is utterly and completely uninteresting to anyone out there beyond statistical correlation with hundreds of millions of other users, and that you can’t realistically expect to hide 90% of your Internet activity anyway. For those times when you really need to duck under cover, though, you should be technically astute enough to abandon those patterns you’ve established.
His logic, which I can’t refute, states that it’s better to have a public persona, which will provide plausible deniability of the few, specific instances when you really need privacy and anonymity, than to constantly attempt to hide everything you do.
Recent Comments