The Three Worst Words In The English Language: Can't We Just?

When learning something new, especially a technical something, it's great to hear the words "for example" because you're about to see something more concrete that helps the abstract make more sense.

Conversely, when doing design and development work, it's awful to hear the words "can't we just" because you're about to hear a defense of kicking the can down the road -- more status quo.

"Can't we just" is used to justify all manner of things:

"Can't we just replicate the passwords?"

"Can't we just leave passwords cleartext?"

"Can't we just use one way SSL to solve everything?"

"Can't we just use the same system we have for the last 15 years for our new Cloud?"

"Can't we just use the same system we have for the last 15 years for our new Mobile apps?"

"Can't we just hardcode XYZ?"

What people are really saying when they say "can't we just" is, "Can't we assume tomorrow will look like today?" This may work in some areas of IT (although I am doubtful), but it's flat-out hazardous in security.

The prefix for the majority of suboptimal designs is "can't we just." When you hear that phrase, brace yourself. I am all for being practical, but systems age more like milk than wine. They do not necessarily get better with age, software evolves, and, more importantly, so do attacker's capabilities. To illustrate this latter point, consider Dave Aitel's prediction for 2012 that mobile platforms would fail:

You know what didn't pan out? "Mobile attacks" in commercial attack frameworks. The reasons are a bit non-obvious, but deep down, writing Android exploits is fairly hard. Not because the exploit itself is hard, but because testing your exploit on every phone is a nightmare. There's literally thousands of them, and they're all slightly different. So even if you know your exploit is solid as a rock, it's hard to say that you tested it on whatever strange phone your customer happens to have around.

And of course, iOS is its own hard nut to crack. It's a moving monolithic target, and Apple is highly incentivized by pirates to keep it secure. So if you have something that works in a commercial package, Apple will patch it the next day, and all your hard work is mostly wasted.

Now there are some good reasons why this did not happen; in my view, the fragmentation in mobile is a real issue for developers, testers, and attackers. But it's not a long-term advantage -- it's a delay of game, a speed bump. The point I would like to make here is that you could read those comments as, "Well, it's too hard for attackers. Can't we just live with the same mobile security model in 2013? After all, it was good enough in 2012."

But guess what? It's not a static environment. Attackers don't remain in some little, limited snow-globe world: They learn, and tools get better. What was good enough last year is not good enough this year.

Instead of can't we just kick the can down the road, we should find a way to make improvements in our security architecture.

Gunnar Peterson is a Managing Principal at Arctec Group