-----BEGIN PGP SIGNED MESSAGE-----
Hash: SHA1

I'm working on a open source project using Crypto++, the project will
be supported on both LINUX and WINDOWS. The concept behind my project
is to improve the quality of future Anti Virus and Intrusion Detection
products by testing our current state of detection mechanisms such as
heuristics, emulation, signatures and reverse code engineering against
malicious software samples using cryptology to evade detection. My
product is in its very beginning stages and I have a lot of work to do
but progress has been made and the theoretical is possible. I believe
in our current state we are not prepared for Cryptovirology. I am
currently not aware of any AV vendors who support any true methods of
detecting cryptographic malware. I envision a future where Anti Virus
companies will need to start thinking of new ways to detect malware
using cryptography and how to protect customers. There are plenty of
AV vendors dominating the market with great heuristics and advanced
methods of taking snapshots of memory and spotting malicious behavior
through emulation and debugging. Sure once the file has decrypted
itself in memory or on disk but what if the file has not been defined?
How do you define something encrypted/self decrypting files as being
malicious in the first place? If the file is obfuscated, encrypted and
signed with a stub binary for instance, handling the decryption,
verification and execution do you flag the stub at the EOF or do you
flag the signature of the sample (say for instance the malware binary
was signed using a public key). I think the first step is to prove AV
software, in it's current state is no challenge against cryptography
or even detecting a valid RSA key with a bit length of 1024, 2048,
4096 embedded inside say the .rsrc section of an EXE. However, if
vendors start thinking of methods for verification of encrypted files
and then design a method to implement a concept to detect stubs
containing functions that verify, decrypt, unpack and execute payload
we may have a defense against it. I also believe vendors need to
"define" these viruses, trojans, worms, bots,  rootkits, malice BHOs,
spyware, adware, shellcode etc... that are using cryptology with
something that identifies them as being encrypted in the first place.
A file named infected.exe for instance could be defined as variant
foo.rsa.xyz or foo.aes.rsa.signed.xyz variant. I don't think it's
ethical for anti virus to not disclose to it's customers that a file
may possibly contain cryptographic functions which could be considered
harmful to their machine, the fact that having a file like foo.rsa.xyz
containing high grade crypto could actually be a violation of export
outside of the United States of America. Sure, one might argue that
they didn't have any idea variant foo.rsa.xyz contained crypto that
under certain laws requires an export license before exporting it
outside of the US but are there laws protecting the unknowing victim
from this? I have many questions but the ones that concern me the most
are related to the failure in our current detection techniques in
regards to cryptographic malware. We have a hard enough time as it is
defeating anti-debugging tricks, compression from various packers,
code obfuscation, etc...  I will summarize my  statements  with this,
will we wait until it's to late? What will they do when PGP for
rootkits is released? It's time these vendors who we put our trust in
start working on the issue. Thank you to all who have taken the time
to read this and to all who respond.

Thanks Wei for Crypto++ I hope to do something useful with your library.

Regards,

Dillon Beresford
-----BEGIN PGP SIGNATURE-----
Version: GnuPG v1.4.9 (GNU/Linux)
Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org

iEYEARECAAYFAkl4jF8ACgkQRnxC5lZRuuFqqACbBtBNWhZSJwisOsJgL+uVaDxy
yrsAn1KagFVEwlgWBW3B+IbNeQ8/GYx6
=XA6w
-----END PGP SIGNATURE-----


--~--~---------~--~----~------------~-------~--~----~
You received this message because you are subscribed to the "Crypto++ Users" 
Google Group.
To unsubscribe, send an email to [email protected].
More information about Crypto++ and this group is available at 
http://www.cryptopp.com.
-~----------~----~----~----~------~----~------~--~---

Attachment: 0x5651BAE1.asc
Description: application/pgp-keys

Reply via email to