I'm not a cryptographer by any stretch, but I have a thought on this. What if we used an asymetric method where the system submits a random string to the dongle, the dongle signs that string and returns it, and then the system verifies that the signature is valid. I don't see any way that reverse engineering the software would let us figure out what the correct response is because the system doesn't do what the dongle does, it only verifies that the dongle did it correctly. It still doesn't solve all the problems, but it makes it harder to crack it because the security lies only in keeping the keys safe (assuming the implementation is good), which is what we're supposed to be striving for... Just a thought anyways... jordan ----- Original Message ----- From: "c0ncept" <c0nceptat_private> To: <PEN-TESTat_private> Cc: "Ryan Permeh" <ryanat_private> Sent: Tuesday, June 05, 2001 3:36 PM Subject: RE: How secure are dongles for copy-protection? > > If the dongle performs some cryptographic operation based on constant > information from the machine, then it should be possible to record and play > back the information. > If the information is based on a psuedo-random value, then some way must > also exist to verify the authenticity of the returned value on the computer. > The application would have to perform the same operations in software, and > compare the result against the value it computes. > A similar hardware example would be the RSA SecureID / ACE Server. The ID > generates a new ID every 60 seconds; the ACE server generates the same ID. A > dongle scheme based on some cryptographic operation would be functionally > simaller, with the caveat that the ACE Server is implemented in software on > the box [ The application requiring the dongle]. Simply reverse engineering > the software would be sufficient. > I may be missing something, but I can't think of a way to implement > something in hardware and have it verified by software in a secure manner > when a malicious user has access to the software. > > --c0ncept > > > -----Original Message----- > From: Ryan Permeh [mailto:ryanat_private] > Sent: Tuesday, June 05, 2001 11:24 AM > To: Felix Huber; Penetration Testers > Subject: Re: How secure are dongles for copy-protection? > > > the only types of dongle protection that don't completely suck are those > that take information from the machine and perform a specific set of > operations on the dongle(prefereably a cryptographic operation, a hash or > crypte/decrypt) purely in hardware on the dongle. This means that the > cracker either has to reverse the entire crypto algorithm(using black box > techniques like known plaintext attacks), including finding the keyed value > on the dongle, or use a hardware lab to actually reverse the hardware. This > has been doen a few times, by both academic groups and security groups like > atstake. you don't only want to look at dongle research, but also smartcard > research and all of the other hardware authentication methods since most of > the techniques to authenticate a user using a fob and authenticating > software using a fob are basically the same. > > This basically works down to a time/money scheme, just like most crypto > adversary equations. Is the data you are protecting(you program in this > sense) worth the value of spending time and money to protect in this manner? > will an adversary think it valueable enough to do the work to break this > scheme? This is all assuming a "perfect" implementation, of course, where > breaking the algorithm/key on the dongle is the easiest way in, and not just > subverting control of the application. Also, keep in mind that if you use > off the shelf componenets for your dongle/algorithm, someone spending the > time and money to crack yours may in turn crack all derivitive and > corrolated products, making the worth of cracking your scheme potentially > more valuable than just cracking your software(ie, look at the decss > scenario. one implementation fell, and with that information, all soon > fell). > > > Signed, > Ryan Permeh > eEye Digital Security Team > http://www.eEye.com/Retina -Network Security Scanner > http://www.eEye.com/Iris -Network Traffic Analyzer > > ----- Original Message ----- > From: "Felix Huber" <huberfelixat_private> > To: "Penetration Testers" <PEN-TESTat_private> > Sent: Tuesday, June 05, 2001 4:05 AM > Subject: Re: How secure are dongles for copy-protection? > > > > Hi, > > > > of course - the most dongle checks were cracked. I have seen 3DSMax and > > other... For more information: > > http://www.google.com/search?q=3Ddongle+cracked > > > > > > > > Regards, > > Felix Huber > > > > > > ------------------------------------------------------- > > Felix Huber, Web Application Programmer, Webtopia > > Guendlinger Str.2, 79241 Ihringen - Germany > > huberfelixat_private (07668) 951 156 (phone) > > http://www.webtopia.de (07668) 951 157 (fax) > > (01792) 205 724 (mobile) > > ------------------------------------------------------- > > ----- Original Message -----=20 > > From: Harold Thimm=20 > > To: pen-testat_private=20 > > Sent: Monday, June 04, 2001 9:43 PM > > Subject: How secure are dongles for copy-protection? > > > > > > I'm looking for any information on incorporating dongles into a = > > software package for copy protection. In particular, I'm looking for = > > information on the Rainbow Technologies Sentinel, but advice on = > > dongle-based copy protection in general is appreciated. > > > > How easy/difficult is it to break this kind of copy-protection? Are = > > there any known weaknesses in the dongle-type systems themselves (as = > > opposed to implementation weaknesses?)=20 > > > > Are there any dongle-based protection schemes that have been cracked, = > > and if so, how?=20 > > > > > > > > (A pointer to a URL would be appreciated, if you have it.) > > > > Thanks in advance. > > > > HAL > > > > > > > >
This archive was generated by hypermail 2b30 : Wed Jun 06 2001 - 13:02:35 PDT