Digital Village

have predicted this without ever seeing a computer.

My final example came to my attention in the past few weeks. MIFARE is a proprietary encryption technique for RFID (radio frequency identification) developed by Philips and Siemens in

is possible to discern patterns in the challenge-response authentication procedure that can be used in a replay attack, and from there it is possible to recover the key from the value of the unique identifier and the observed behavior of the shift register in the authentication

URL PEARLS

When it comes to digital security systems, secrecy is indeed the mother of dysfunction. The security vulnerabilities described in this column were real and betray only the slightest hint of literary hyperbole. For those interested in the details, two of the security-through-obscurity examples were covered in previous columns: the Code Red Worm was discussed in December 2001; and Wireless Infidelity appeared in December 2004 and again in August 2005. The RFID MIFARE exploit was presented at the 24th Chaos Communication Congress last December (see events.ccc.de/congress/2007); a video of the presentation by Karsten Nohl and Henryk Plotz is available at video.google.com/videoplay?docid= 425236768097439- 6650&hl=en. ^c

the late 1990s. MIFARE is an attempt to cryptographically secure the now-ubiquitous RFID space that relies on RF transmission for communication between transmitter and receiver.

process. We’ll create STO category III for this MIFARE vulnerability: turning chip designers loose with CAD/CAM software without adequate education and training.

Following the common theme, the security of the proprietary MIFARE system is predicated on the belief that no one will discover how it works. And, as one might predict, some MIFARE circuits were reverse engineered down to the gate level. The result was the discovery that the random number generation that drove the encryption resulted from a 16-bit key linear feedback shift register based on a master key and a time signature. With RFID sniffing via an open PICC (proximity integrated contactless chip) card and a logic analyzer, it

FAITH-BASED SECURITY

Examples of failed STO could fill a weighty tome. I’ve mentioned three. These examples highlight the consequences of building deficiencies into the design of things or at least unwittingly including them. The flaws would likely have been detected and reported had the code, system, or chipset been carefully analyzed during impartial peer review by qualified professionals.

But I don’t want to leave this critical view of deficiencies at the feet of naive SID or STO. I’m looking for first principles here.

I’ll refer to the common ele-

ments between them under the general rubric of faith-based security—in the most secular sense of this popular phrase. The only thing these two security models have going for them is the unsupportable and unjustified faith that they are reliable. These are manifestations of the technologist succumbing to the self-deception that secrecy and tight lips will cover all design misjudgments.

I propose that faith-based security enter our vocabulary as the default model of IT security. Let’s get the faith-based orientation of naive security in depth and STO up front where it belongs. Think of the advantages. If an auditor asks why we decided to place our Web server on the inside of our enterprise firewall, we report that we have faith in our Internet comrades. Faith is a predicate of propositional attitude, like belief, want, and desire. If someone says they have faith in something, one can’t say “No you don’t,” at least not until someone comes up with a method to read thoughts. The auditor doesn’t have faith, we do have faith; half-empty, half-full. You get the idea.

Since the integrity of a faith-based security implementation is by definition taken on faith, we hold the position that whatever policies and procedures discovered by an auditor were actually intended. So what if our corporate mailer is running on an operating system that hasn’t been supported since perestroika—we have faith in good old “digital iron.” After all, when was the last time you read about some hacker