I'll be speaking at Shmoocon in Washington, DC (February 15-17, 2013). Here is my talk info:
Title: These Go to Eleven: When the Law Goes Too Far
Abstract: Agents line up at a non-descript door which houses command and control servers for a criminal botnet. They enter with a court order to shut it down and seize the servers--but these agents aren't law enforcement--they're from Microsoft. Somewhere across the ocean, a domain goes offline. Customers--some legit, some not--lose their work product. They are on the receiving end of civil asset forfeiture by ICE or DHS which seized the domain with no warning and scant due process. These are but two of the ways the law is pushing its own boundaries to go on the offensive to deal with sophisticated and complex cybercrime. It's time we stop and pause to review these tactics--the results might just surprise you.
The conference schedule is available here.
Lawyer ⚖️, Historian, Navy vet ✈️, Philly and Penn State sports fanatic 🏈⚾🏀, Dad and Husband. Blogging at the intersection of state power and civil liberties.
Showing posts with label microsoft. Show all posts
Showing posts with label microsoft. Show all posts
15 December 2012
18 March 2012
MS12-020, exploits, and words mean things
On Friday, Core Security announced that they had released an exploit for CVE-2012-0002 (patched by Microsoft as MS12-020):
I thought this was great news. I am currently working on a client engagement with a lot of RDP on their network and we use Core as one of our tools. Time to pwn, right?
Not quite.
As subsequent tweets made more clear, the Core exploit was a denial of service module, not code execution. This got me thinking--what does it mean when you call something an exploit? I am a moderately-experienced penetration tester, and I think there's probably a misunderstanding about what this means. To me, an exploit is something that results in arbitrary code execution. Maybe a denial of service that results in a BSoD is an exploit, technically, but that's not what I expect the word to mean.
I hadn't thought about it much this weekend until HD Moore of Rapid 7/Metasploit summed up my feelings earlier today:
So, what do you think?
Is a denial of service module an exploit? Does it really matter, or are we splitting hairs?
EDITED TO ADD: And to add to HD's tweet: what exactly does commercial grade mean in reference to a DoS module? Does it give you an extra special BSoD? It seems the words in Core's original tweet are specifically designed as marketing pull, and as a result, imply that the module is a code execution exploit when it is not. The tweet clarifications made it obvious to me that others were as confused as I was.
Lastly, "Enjoy!" seems to suggest that we will like the result. I don't know about other penetration testers, but I can't think of any clients who have ever asked me to run DoS modules against their systems, and even if they asked me to do it, I'd probably advise them against it. What about you?
I thought this was great news. I am currently working on a client engagement with a lot of RDP on their network and we use Core as one of our tools. Time to pwn, right?
Not quite.
As subsequent tweets made more clear, the Core exploit was a denial of service module, not code execution. This got me thinking--what does it mean when you call something an exploit? I am a moderately-experienced penetration tester, and I think there's probably a misunderstanding about what this means. To me, an exploit is something that results in arbitrary code execution. Maybe a denial of service that results in a BSoD is an exploit, technically, but that's not what I expect the word to mean.
I hadn't thought about it much this weekend until HD Moore of Rapid 7/Metasploit summed up my feelings earlier today:
So, what do you think?
Is a denial of service module an exploit? Does it really matter, or are we splitting hairs?
EDITED TO ADD: And to add to HD's tweet: what exactly does commercial grade mean in reference to a DoS module? Does it give you an extra special BSoD? It seems the words in Core's original tweet are specifically designed as marketing pull, and as a result, imply that the module is a code execution exploit when it is not. The tweet clarifications made it obvious to me that others were as confused as I was.
Lastly, "Enjoy!" seems to suggest that we will like the result. I don't know about other penetration testers, but I can't think of any clients who have ever asked me to run DoS modules against their systems, and even if they asked me to do it, I'd probably advise them against it. What about you?
Subscribe to:
Posts (Atom)

