Compliance vs Security

Almost as exciting as a few other epic throwdowns, I am lecturing tonight for the University of Washington's infosec certificate program. A few quick highlights from my lecture notes, which is based on the Source talk I gave this summer.

Compliance-driven security forces you to make certain bets on the big enterprise roulette table - but I only have so many chips to play, so I prefer not to be constrained in my choices.

As a consultant I saw primarily two kinds of organizations:  Those practicing good risk management who wanted to get better and  Those forced to be more secure because of compliance or a breach.

Why is there such restrictive compliance regimens? Without repeatable, evidence-based, agreed-upon risk methodologies, you cannot rely on third-parties to make security decisions with your data that are aligned with your interests, instead of theirs.

Compliance is a multi-dimensional object... and lot more than three.  You've got width - the general rules f the standard plus a few specific new ones based on how the organization interprets it.  This is the easiest dimension.  Depth: As most compliance acceptance is based on auditor opinion, which is driven by the individuals experience.  Plus, if the standard is somewhat worthwhile, it includes the appropriateness of risk model (relevance) to your problem.   Then there's several dimensions of scope: Time (past events, present controls, future possible events) and then the general usual dimension of Physical, virtual, sofware, network... what’s constitutes a a barrier in those domains.    And of course, all of this is moving.

Security is also multi-dimensional but it has slightly different dimensions and moves differently than compliance.

 Best practices?  In other words, “This worked in our organization once upon a time, So it should work for you too.”

Where I live is the intersection of:

1. What the auditors demand we do,

2. What we need to do to keep from getting breached, and 

3. What we can afford to do.

And I'm mot going to get all of all three.

Stupid compliance failures:

- Why is the absence of a particular control is a risk? A high risk?

- How can I be 100% compliant with an open standard? With a product lifecycle of 12-18 months?

- Hey, that's a feature not a high-risk vulnerability - it all depends on your context

- Impact does not equal risk. You forgot probability. Dumbass.

I do it for the Lulz

I've always done for the lulz.   Security that is.  When I lecture to students up at UW, I try to warn them that don't do this job for money. 

Anyone going into infosec for the money, the prestige, or for the job security is seriously misguided.   I tell them the only reason I do it is for the lulz.  I warn them to prepare to face the humiliation of having a server hacked, the terror of knowing the bad guys will outspend and outlast you, the tedium of when nothing happens and the crunch when you need to justify every thing you've done to an auditor and the budget axe.  It's tough, it sucks, it's relentless, and I still love it.

What are my lulz?  The thrill of the chase. Heck, forget the chase, how about actually taking down some bad guys?

Besides the sexy stuff, I also get lulz making an organization safer… or even if it's just a friend of the family that needs some malware scraped off their machine.  Sure, it's tough work but it feels good to make the world a little safer, a little saner when you're done.  And knowing that you've deprived some creep one less victim.   Lulz.

I get my lulz designing new systems, making the strong, making them resilient, making them better than they were before.  And digging deep and figuring out where the holes are, where's the best place to fix things, and then working on presenting that to the people that care.  Even more fun than all the puzzles and video games in the world.


Technology, and especially information security has always been more than a job to me.  More than even a career.  It's a calling.  Don't tell my boss, but I'd do this even if they didn't pay me.  It's what I do.  I can't help it. 

And to those who say we're losing the war.  Whatever.  I've been hearing that for years.  The world hasn't ended.   I know that more systems than ever are online now and somehow we failures are still protecting a majority of them.   I know we'll be always outnumbered always outgunned.

That's what makes it a challenge.

Decompiling the week.

What an amazing week...

Fantastic time at Source Seattle.  If you didn't make it, you should really check out what you missed.

Great keynotes by Kris Herrin and Eric Cowperthwaite.  Nice getting the executive "big picture" on breaches and managing security.

Thoroughly enjoyed giving my talk and a lively audience as well.

Fascinating lunchtime discussion with Marcia Hofmann about privacy and the nature of social media.  Enough to make me re-up my membership in the EFF.  You should consider it too.

Not only a great demo by Ron Gula, but he spent time after the session doing a one-on-one with me giving an insider's tour of their software.  It was great to see a master at work.  How often do you get that kind of access to that caliber of talent?

If Source wasn't enough, I had to get me some Agora where Kirk B. pointed me at this fascinating paper on assuming a state of compromise.  Since that's what this blog is all about, you should check it out.

Now I need to sleep...

The Kobayashi Maru

Trek nerds will remember the Kobayashi Maru as a requisite test for command.  It was a simulation of a no win scenario that taught a candidate would deal with utter failure.  As Spock said, "The purpose is to experience fear, fear in the face of certain death, to accept that fear, and maintain control of oneself and one's crew. This is the quality expected in every Starfleet captain."

I'll also say that this is a quality I expect in every security leader.  Except our fear isn't death, but of breach.   Like the title of this blog, I think it's a useful exercise to assume you've been breached and plan accordingly.   For some, this is as radical idea of contemplating one's own mortality.  Specifically, I've encountered more than a few executives and tech leads who are fully willing to go their entire career expecting that they will never experience a data breach.  For me I saw as an educational opportunity to teach.   Teach them that organizations can survive a breach; it's a matter of doing the best job you can and being able to prove it.  It's also a matter of knowing where your weak spots are and what can happen.  And it's a matter of preparing for response.

If you take nothing else out of this post, take this:  perform a Kobayashi Maru test on yourself.  Test your incident response plan.   There are some great guides out there. Write a plan and test it.  Figure out some likely scenarios and run the steps and see how you do.  For scenrios, you could even replay the last few major breaches and think about how you'd do if it happened in your org.  Not how you'd defend against it (cuz I assume y'all thought of that the second you read about it) but imagine it already happened.  Now think about how far it's spread internally, what data would be leaked out, what services would be offline, what forensic data would you have?  This will likely cause you rethink some controls - are you logging enough?  Do you really have defense-in-depth?  Do you have an accurate data inventory?  Do you have all the critical personnel on speed-dial?  Do you have an organized method of contacting customers?  Figure this all out and share the data with your boss.  Tell her it's a good idea to plan for a disaster so it doesn't destroy the company.  How an organization responds to a breach is crucial factor in a security program. 

And when you want to point at other major breaches and chuckle with schadenfreude, you should think one thing - that could have been you.   You think you've got all your bases covered, you're locked down and unbreakable?  Think again.   And you know what, check again on those companies in 12 months and see how they're doing. Some are done and gone.  Others have survived, maybe even stronger.   And to those security folks there, I think they might have done a good job preparing for failure.   And I take that as a challenge. Again, Trek said it better than me.  This time it was Captain Pike, talking about the destruction of the USS Kelvin. "Your father was Captain of a Starship for 12 minutes. He saved 800 lives. Including your mother's and yours. I dare you to do better. "

What I've learned about Rugged in the past 24 hours

Well, I learned yesterday's post touched something in some people.  Based on the comments I got, both online and offline, I can guess a few of us are confused about what Rugged is about.   Especially those of us who've only read about it, instead of having it explained to us.   And sadly, some of us "saw it and dismissed it" as another security fad. Maybe this post can help fix that.

A lot of folks, including @joshcorman himself, stepped up to help me understand Rugged. Very nice, Lazyweb!


First, let's start with the problem (as I see it)


Software security programs have a poor Raison d'être. This is likely because it's hard to define what "secure software" is.  (heck, define "secure")  Is secure software?

• Resistant to cross-scripting and SQL injection attack (insert attack du jour)
• Bug-free?
• 100% OWASP complaint (yes, I have been asked this)
• Have no high vulnerabilities?
• Made with high quality?

Waste of time, right?  We all know secure is a sliding scale based on value and risk.  You can't arbitrarily define security, which makes it less than useful for talking to executives and business program managers.  So how do we frame the conversation in a useful manner?  Enter Rugged.

Rugged leap-frogs over all these definitions and points to the qualia we security grognards are jumping up and down about.  It brings it down to earth with a clear and sharp image that conveys the essential intrinsic properties of "secure software"

To answer my own questions:

1) How is Rugged different than any other Best Practices?
Well, it's NOT really a best practice… more of a framing technique… or (ulp) a paradigm.  I was expecting too much of Rugged to even put it in this category, it's just not that kinda thing.  It's just a way to simplify the dynamic and intangible.  Of course, we could apply some evidence-based analysis over time to see how effective it is in helping the non-security folks understand us.

2) Convincing the developers to write more secure/stable software isn't my problem. My problem is convincing customers and managers so that they'll let/encourage the programmers to to write more secure code.
Ah, this would be Rugged's sweet spot.  Here is a meeting ground for the security team, developers and money spenders to agree on something that is useful and clear.   A way to communicate what needs to be asked for, what needs to be done and what the final product looks like.

3) Software security problems are deep and complex.
Actually, digging deep enough into Rugged, this issue is acknowledged.  And Rugged doesn't aim to solve these problems directly, but again, it gives us all something we can put hands around when wrestling with them.

4) Rugged appears mysterious and embryonic.
Hopefully we can change that.  The more we spread the word (and ask questions), the less confusion we'll see.  So I'll light a candle now:

Here's how I would summarize it as guidance from management to the developers.

"If our software is Rugged, it is built to withstand adversity, tolerate anomalies, and always do what we intend it to do. Our customers depend upon this level of unyielding reliability; in fact, they expect nothing less. It is our responsibility to meet these expectations."

How's that sound?

Would someone please explain this Rugged thing to me?

I'm steeped in a huge SSDL project here at work - looking to move security in our development processes to the next level.  Lots of heavy lifting doing evaluation, analysis and reorganizing.  I'll throw in a shameless plug for WhiteHat Security who's helping us a ton.

Now, one of the things that came up in my search to see how to improve things was the Rugged Software movement.   Early on in the process, I foolishly mentioned it to our CTO as something to look at.   Why did I saw this was this foolish? Well, because at the time, I had only a cursory understanding of Rugged.   He went off and dutifully checked into Rugged only to find the bare documents on the website.  Indeed, it was a movement, but apparently not much else.. at least at that stage of the game.  He came back to me confused and wondering why I had brought it up to him.  What was he supposed to do with this Rugged thing?  Oops, I had just wasted some credibility and an important ally's time.   A mistake I wasn't going to repeat.

Well, here we are months later, and I'm afraid I still have only a cursory understand of Rugged.  Apologies to Josh and the other creators of Rugged, but I just don't see anything there worth passing on yet.  Maybe it isn't aimed at our developers? I don't know.  It wasn't clear.  I'll be the first to admit I've not attended any conference talks on Rugged (I admit here, I don't make to many conferences) and I don't attend many webinars or online thingies (they're often hard to follow).  I have googliated a bit and haven't found much beyond a few news articles.  On the other hand, I have found tons of advice and guidance on practical secure development frameworks like BSIMM.

Overall, my big questions / confusions are:

1) How is Rugged different than any other "Best Practice"? 
Is there any evidence yet to show that it improves security?  Can I see it?   Can I share it with management?


2) Convincing the Developers to write more secure/stable software isn't my problem.
Talk to them, as I have, and most of them wouldn't mind writing more secure code.  Some of them even want to write more secure code.  And a certain chunk of them don't know how to write secure code.  I don't see how Rugged solves any of these problems very well.    The root of the problem comes from the fact that secure code still isn't spelled out in the requirements.  Developers can only do what the project manager demands, which is based on what the customer demands.   So if Rugged is aimed at convincing customers to ask for more rugged software, specifically and pointedly asking, then I'll admit it should be preached (but not to me, to my customers).


3) Software security problems are deep and complex.
A lot of security bugs are buried deep in old crufty code or libraries.  Even when all our developers are cracking on all cylinders of secure code dev, we're still excavating for fundamental faults and design flaws.  And when you land in those pits, you're dealing with Expensive Questions - redesign ($$$) or patch-and-move-on.   I need a movement that helps me make those decisions.


4) Rugged appears mysterious and embrionic.
I'm sure it will grow up to be influential and useful, but to be practical to me right now, I need something that's actionable that I can use with my Development team and management.    The story I mentioned in the opening about confusing my CTO cannot be repeated.   And beyond that, my executive team will ask for proof and metrics for any new development movements I propose.   I don't blame them.


So please, help me out here.  I am confused, what am I missing or misunderstanding?

UPDATE - People have stepped up to 'splain it to me (ha, my evil plan worked).  Read what I've learned here.

Good pen test reporting resource

I knew there were folks out there who could do a good job at this.   Instead of writing sloppy security reports, here's a positive example of how to do a better job at it by Steve Shead.  He's a security guy and a graphic designer, so no wonder I like his layout for pen test reporting.