Can you trust 'NSA-proof' TrueCrypt? Cough up some dough and find out
Fundraising drive to shove a microscope up mysterious encryption software
Security researchers are raising funds to conduct an independent audit of TrueCrypt, the popular disk encryption utility.
TrueCrypt is widely used as a tool to strongly encrypt and decrypt entire drives, partitions or files in a virtual disk. It can also hide volumes of data and said to be easy to use.
The source code for the Windows, Linux and Mac OS X utility is publicly available for people to inspect and verify, but this has not been enough to convince every cryptography guru that it's entirely secure.
For one thing, researchers have been unable to prove that the downloadable Windows executable, built by the TrueCrypt team, can be constructed purely from the published source code, for reasons based on unusual decisions by the developers – as explained by cryptographer Matthew Green here. (In short, the Windows binary appears to save a block of unexplained bytes with the encrypted data. Some fear this is a key to a backdoor, which would allow people in-the-know to decrypt the data without the user's password.)
Encryption authority Bruce Schneier has recommended TrueCrypt as a tool to keep sensitive files out of the grasp of the NSA's global data dragnet, albeit it with caveats. He stops short of giving it a ringing endorsement.
"No, I don't have any inside knowledge about TrueCrypt, and there's a lot about it that makes me suspicious," said Schneier. "But for Windows full-disk encryption it's [TrueCrypt], Microsoft's BitLocker, or Symantec's PGPDisk - and I am more worried about large US corporations being pressured by the NSA than I am about TrueCrypt."
TrueCrypt's documentation makes it plain that it can't secure data on a computer compromised by malware or a hardware keylogger. It's also well known in computer forensics circles that TrueCrypt keys can be recovered from memory, even using commercial tools from the likes of ElcomSoft, given physical access to a powered-up machine. So-called cold-boot attacks allow the same trick to be tried on recently powered-down devices.
Encryption tools are not a panacea. Unless a user follows best practices and operational security guidelines then their precautions will be stripped away by cops, the Feds, intelligence agencies or other capable attackers.
Hidden backdoors? Who knows
That's always going to be the case with any security or encryption tool but the concern here is that TrueCrypt may be unsafe even when it's used properly because of a hidden backdoor or similar. These concerns have always been present, but have risen to the fore because of the ongoing controversy over Bullrun, the NSA's effort to work with hardware and software technology vendors to weaken encryption systems and their underlying components.
A new project aims to crowd-source funding to audit TrueCrypt and compare its published source to the compiled binaries in circulation. Such an audit is long overdue, as the security experts who kicked off the fundraising drive explain:
We want to be able to trust it, but a fully audited, independently verified repository and software distribution would make us feel better about trusting our security to this software. We're pledging this money to sponsor a comprehensive public audit of TrueCrypt.
The project was created by Kenneth White, the principal scientist at biotechnology biz Social & Scientific, a hosted services provider to the health sector, and Matthew Green, a cryptographer and research professor at Johns Hopkins University. The project's goals include conducting a public cryptanalysis and security audit of TrueCrypt version 7.1a, one of the latest builds, as well as sorting out licensing issues that have prevented the suite from being bundled with Linux distributions including Ubuntu, Debian and Red Hat.
In addition, researchers also want to certify a build of the security software for Mac, Windows and Linux users. They hope to apply a variety of approaches and tools to study the code, including potentially paying out bug bounties as well as paying for professional fingertip-searches of the code.
An evaluation [PDF] of the deniability of hidden volumes in TrueCrypt by Schneier and other experts five years ago discovered security leaks and other causes for concern. These issues may well have been fixed by now but without a proper audit we just can't be sure.
Scrutinizing code is painstaking work that can only be carried out thoroughly by experienced and skilled practitioners, so the only way to nail this audit is to ensure it's properly financed. Hobby projects have resulted in improvements in cryptography in other areas but auditing TrueCrypt is a far more serious undertaking that can't be left to amateurs.
The project has attracted $2,922 via 36 pledges to FundFill and a further $1,640 through Indiegogo by Tuesday afternoon; a fair way towards its funding target of $25,000 within two months.
A blog post by Green outlines several reasons why an audit of TrueCrypt is needed and arguably even overdue.
"The 'problem' with TrueCrypt is the same problem we have with any popular security software in the post-September-5 era: we don't know what to trust anymore," he writes. "We have hard evidence that the NSA is tampering with encryption software and hardware, and common sense tells us that NSA is probably not alone. TrueCrypt, as popular and widely trusted as it is, makes a fantastic target for subversion."
Green is careful to say he has no specific reasons for doubting the strength and security of TrueCrypt. The audit is proposed more in the spirit of “trust but verify” than a search to confirm a suspicion. The developers of TrueCrypt are anonymous and this is a major reason that Green et al are uneasy about TrueCrypt, and one of the main reasons it has become the target of a funding drive.
"We don't have an upper limit on Fundfill. I'm talking to audit firms this week," Green told The Register.
TrueCrypt's anonymous developers could not be reached for contact. Their website denies of any suggestion that TrueCrypt has a hidden backdoor. The same statement explains that it's not possible for TrueCrypt to assist in decrypting data in cases where users have forgotten their own password – suggesting once you've lost the key, there's no way of recovering the data. ®