Black Hat Blog There are two rules in Las Vegas. One: everything is twice as big as you think it is and therefore twice as far away. Two: wherever you need to get to is across a casino.
Into this distorted sense of space Black Hat fits nicely: its 3,000 attendees are more IT security people than you might have imagined exist in the world. Surely this many security folks ought to be able to fix everything that goes wrong in our computers every day.
Apparently not. I learn today that although computer forensics has been with us as a discipline for some years and despite many well-shock-horror-publicised break-ins, there is still no commercial OR free tool to do database forensics. The importance of being able to find out what happened in such cases seems unquestionable: last January, the world found out that a database of 45.6 million credit card numbers was stolen from TJX. But the attackers copying the credit card information had intruded on the system as early as January 2005.
But tech and counter-tech. For every person here who's trying to block – or at least trace – intrusions there's someone trying to circumvent the blocking. Later in the day, three speakers from iSEC Partners discuss breaking forensics software. Their purpose is sound: to query whether the trust the courts place in the results of forensic software might be misplaced. Based on a study of EnCase, the authors concluded that present testing of such software focuses on ensuring that it won't miss hidden data rather than more traditional security testing. Files especially configured to crash the system when examined by anything other than the software used to create them are a particular vulnerability.
Other talks examined the vulnerabilities of VOIP which Zfone is trying to overcome. Brought to you by the original author of PGP, Phil Zimmermann, Zfone sounds worth trying if you use any of a number of common VOIP software packages (not, however, Skype, as that service keeps its inner workings secret). Thirteen years ago, we thought all telecommunications would be encrypted by now.
But counter-counter-tech: those who rely on anonymising services might wish they didn't. Ricardo Bettati, from Texas A&M University, has a technique for using timing data to reconstruct the actual data flows from mix networks (these are networks designed to aggregate traffic so that a third party can't tell who's communicating with whom). The upshot is quite sophisticated traffic analysis, the technique of looking only at metadata rather than the content that metadata surrounds. Most privacy laws focus on protecting content; but that, argued Jon Callas, the CTO of PGP Corporation, is protecting the wrong thing.
Mind you, once you've heard Johnny Long do his talk on "no-tech hacking", you'll conclude that no one ever thinks to protect the right things. What Long can find out using ordinary objects like a TV, a camera, a little sleight-of-hand with a piece of plastic, and ordinary observation would make Sherlock Holmes decide he needs another line of work. ®