‘Non-Code Codes’—a Way to Maintain Office E-Mail Privacy?

That's not a valid work email account. Please enter your work email (e.g. you@yourcompany.com)
Please enter your work email
(e.g. you@yourcompany.com)

 The boss of an IT-whiz friend of mine has ordered him to cc or forward to him [the boss], all of his [my friend’s] emails to or from any other staff member, regardless of their content or email service used.

That’s because my friend’s computer is the only one in the office network that is not under real-time, on-screen surveillance with live access to employee emails.

[My friend has refused to comply and is currently looking into new job options for his formidable, high pay-grade talents.]

Now, that’s really intrusive. So, what can you do to maintain some semblance of privacy and confidentiality—not to mention dignity—under such circumstances?

Well, you can’t encrypt your on-the-job emails into indecipherable strings with encryption software. That would merely provoke a confrontation, or at least raise grave suspicions.

Encoding “Hamlet” as “Duck Dynasty”

Instead, consider writing everything in readable, ordinary-language code—seemingly “non-code code”, if you will. If you think that’s not easy, you’re absolutely right, especially if you take things a step further and encode everything you write as perfectly normal, coherent text that doesn’t at all resemble anything coded.

That would be like undetectably encoding and transcribing the entire text of Shakespeare’s “Hamlet” as an episode of “Duck Dynasty”.

I don’t know enough cryptography to even begin to answer the question whether that is humanly or mathematically possible. Sure, individual words, no problem.

For example, take care that, when writing about how many miles you traveled on your recent holiday, you don’t let your entire right-hand slip one character to the left on your keyboard, because then “miles” will appear as “nukes” in your email—with results you can imagine.

So, as this “method” shows, one word or maybe even a sentence or two may easily be possible. But paragraph after paragraph? Here’s an imaginable short reply, at work, to a fictitious emailed question about how many air miles I have on my Royal Bank of Canada Gold Card, using the simple shift-code described above: “Gave 55,356 nukes!”  [“Have 55,356 miles!”]. The point is that even this much is a challenge.

Now, you may think suggesting you’re handing over nukes is worse than revealing how many air miles you have; but, if you think about it, the last thing a boss wants you to be thinking or writing about at work is your travel plans.

Be that as it may, the point is that it wasn’t easy to create even that much “code”, even when half the keyboard—on the left—required no special coding.

Hence, it seems to me that writing paragraph after paragraph to a “Hamlet”-“Duck Dynasty” encoding standard will be impossible for just about everyone, with the possible exception of those whose job is to do precisely that—which, of course, suggests they are probably already doing that with their own office emails as well as with their paid projects.

However, don’t despair. There’s a simple workaround : Just preface every email you send, encoded or not, with “THE FOLLOWING MAY BE AN ENCODED MESSAGE”. If it is intercepted by your boss or his software, you’ve got “plausible deniability”, e.g., “It’s really my project report, which I wanted to protect from unauthorized interception or surveillance. The problem is that my dog ate the code before I could code the code and send it.”

Kolmogorov, One Million Monkeys and Your Boss

This office-email coding issue resonates with the well-known Kolmogorov test for complexity and randomness, e.g., of codes and that which they code, which, among other things, specifies that the longer the sequence of characters required to describe a second string of characters, the more closely the latter approximates a purely random, complex sequence.

For example, both the code for and the sequence consisting of “1234123412341234……”, endlessly repeated, are simple and non-random, the code being “write ‘1234’; repeat indefinitely’”. On the other hand, “15436133221914378574….” [continued with no evident pattern whatsoever] represents a code and a message far more complex and random.

As an upper limit, a purely random, infinitely or indefinitely-long complex coded message’s description and code is equivalent to the message itself. So, on analogy, can it not be argued that as non-random text indefinitely increases in length, the only coded descriptive “recipe” for it becomes the text itself?

[Note: This tandem relationship between increasing complexity and increasing randomness of information needs to be compared with and distinguished from the relationship between biological complexity and randomness, which, on the face of it or at least on some interpretation, seem to vary inversely.]

The challenge of coded transcription of “Hamlet” into “Duck Dynasty” or of your emails into coherent concealment is also a variation on the “million monkeys with a million typewriters” thought experiment, which imagines that given enough time, one million monkeys will indeed flawlessly type out a compilable copy of “Hamlet”.

However, in the email case, the prediction is not that eventually they would write “Hamlet”, but that they would/would not be caught and exposed after writing a “Duck Dynasty” episode as code for “Hamlet”.

To Be or Not to Be Cagey

Therefore, if you preface every in-office or other email with “THE FOLLOWING MAY BE AN ENCODED MESSAGE” or “THE FOLLOWING IS AN ENCODED MESSAGE”, you may be able to retain some measure of privacy, if only because of the confusion you and the attendant employer exasperation will sow.

As for which of the two versions, i.e., the “IS AN ENCODED MESSAGE” or “MAY BE AN ENCODED MESSAGE”, the choice will depend on whether you ever expect to have your coding abilities challenged or tested by your boss.

One advantage of the “MAY BE” version is that it amounts to a randomized strategy that makes predicting whether or not you’ve encoded your emails very difficult.

Besides protecting you from spot tests and challenges, it will, as a minimum will keep your boss off-balance and uncertain as to when to confront you.

Hence, for now, for me and probably for you, “MAY BE” is likely to be the wiser choice.

…unless you think that at least one of us is looking at a perfectly coded article arguing the contrary and that you can prove or match it.

___________

Note: In case it went unnoticed, there was some intended humor in this.

By Michael Moffa