True mystery of the disappearing TrueCrypt disk encryption software

Filed Under: Cryptography, Featured

Webdriver Torso has nothing on this week's mysteries!

First we had Apple iDevices in Australia announcing "Device hacked by Oleg Pliss" and demanding a $50 Moneypak voucher or $100 via PayPal.

No-one seems to know how, or why, and (to make things yet weirder) the PayPal address given for payment didn't actually exist.

Now, the website of venerable free disk encryption software TrueCrypt is telling us:

WARNING: Using TrueCrypt is not secure as it may contain unfixed security issues

Not only that, but the page goes on to state that the project has been closed down, following the end of support for Windows XP:

The motivation seems to be that on all supported operating systems for which TrueCrypt was available, there's now some sort of built-in full disk encryption system.

Yet more curiously, there's a new version numbered 7.2 that can apparently only decrypt, intended to help you migrate away from the now-defunct TrueCrypt product.

The 7.2 "decrypt only" version can be downloaded for Windows, OS X and Linux.

What gives?

Is it a hack?

If so, no-one from TrueCrypt has yet come forward to suggest so.

Is it malware?

We don't yet know, but the unlikely, unprofessional-looking and abrupt download page suggests that you'd be unwise to trust anything about it.

Is it the result of a legal threat?

Some have suggested that there might be legal pressure on TrueCrypt, combined with some sort of gag order that prevents the full story being told, so that this is the only lawfully clean way of closing the project.

Sort of like what happened to Lavabit, but deliberately wrapped in mystery.

But wouldn't just closing the project with no explanation at all be even cleaner?

Is it a panicked response to a failed audit?

TrueCrypt announced a big code audit recently, as a way of restoring confidence in encryption software following the many Snowden allegations about government surveillance.

The results of the first part of the two-stage audit were published last month, apparently finding nothing untoward at all.

But even if the second part of the audit had revealed something terrible, why not simply say so, given that the audit would be published anyway?

Is is a side-effect of SourceForge's recent forced password reset?

SourceForge recently announced it was boosting the security of its password storage, and would require everyone to choose new passwords as part of the improvement.

But it's not clear how this could lead to a TrueCrypt shutdown.

Even if crooks had acquired TrueCrypt's SourceForge passwords in the past, and realised their window of opportunity to abuse those passwords was closing, why take this sort of action now?

Is it a publicity stunt?

Well, if it is, it's worked, but not in any positive way.

The bottom line

It certainly looks as though TrueCrypt is finished.

If the new web page is true, the project has ended explicitly.

If it isn't true, then it's going to be tough to re-establish trust in the code, and the project has ended implicitly.

→ The mysterious decrypt-only version 7.2 is apparently digitally signed with TrueCrypt's private key, making the most likely explanation either that this is intentional by the developers, or that it is an unlawful hack. TrueCrypt's private key would not be accessible via a SourceForge compromise, implying that this goes deeper than the Sourceforge-hosted content.

Do you know anything we don't? Please tell us if you do - we'd love to get to the bottom of this mystery.

, , ,

You might like

70 Responses to True mystery of the disappearing TrueCrypt disk encryption software

  1. Anonymous · 61 days ago

    wow, that is surprising news indeed, and unwelcome! Next thing, eraser will disappear. Are there any alternatives? It hope the open source security community starts another data encryption project. I use true crypt daily.

    • wolfenpl · 59 days ago

      Source code version 7.1a is in a lots of place. Very easy will be take create fork of truecrypt. Open Source is very hard to kill. Probebly few people right now working to change this version TC to something new. Only for start will be hard to trust new software.

  2. John W Baxter · 61 days ago

    I know nothing. However, my initial reading go the note at the top stands (for me): we're no longer working on this and while we know of no security issues, there may be some. If there are, we would not fix them, so don't trust the old code.

    • Paul Ducklin · 61 days ago

      Why not say, "We've given up working on this, and we haven't got all the known bugs out, so we suggest switching to some other product"?

      Or do you think that the "we're no longer working on this" might be because of a rift in the coding team, causing the abrupt end of the project?

      • John W Baxter · 61 days ago

        Paul, your suggested wording is better than mine, but I read it to have the same meaning. We'll probably find out sometime.

  3. Unknown · 61 days ago

    According to Wikipedia.com, one of the planned features according to the TrueCrypt website, was "Full support for Windows 8".

    Reference:
    http://en.wikipedia.org/wiki/TrueCrypt#Planned_features

    • LindaB · 61 days ago

      Wikipedia cannot be trusted as a valid and reliable source. Too many entries are incorrect. A recent 'audit showed some 82% of pages about medical conditions were either misleading or plain wrong!

      • Please provide a link to the audit.

      • Anon · 60 days ago

        Just because Wikipedia has a few errors on medical articles, doesn't mean it can't be used - besides, what do technical medical articles have in common with articles on encryption software? If you actually look at the statistics, classic encyclopedias have higher error rates than Wikipedia. Furthermore, as there is no current contact with the TrueCrypt developers, Wikipedia is one of the best sources.

      • 73.6% of all statistics are made up.

      • Paul Ducklin · 60 days ago

        In this case, it seems that the original commenter is on safe ground, as the info about Windows 8 was almost certainly just copied-and-pasted from TrueCrypt's own website.

        (TrueCrypt didn't show any previous sign of imploding, and "full support for Windows 8" would be a sine qua non for continuing to develop it anyway.)

        • Hearth · 60 days ago

          This is the only reason I feel even slightly suspicious of the shutdown, the abruptness of it all. Although, we have had no activity on the project for over 2 years, so maybe the Dev(s) just don't have time. 10 years ago I had no family and was young enough to spend all-nighters working on open source projects - this is no longer the case.

          I don't think there is anything wrong with the software and I believe the audit will show that (with the exception of minor coding issues as we have already seen). If it wasn't for how quick-and-dirty all this went down we probably wouldn't even be talking about NSL's etc. but I have to admit there does seem to be something "fishy" going on.

          I hope the project is picked up by some FOSS group as it would be a great loss to the wider internet community. It would be great if the devs were to publish the forum/knowledgebase that has gone with their website as well - someone else mentioned this somewhere - all that knowledge and hours of community work would be valuable if not just for posterity.

      • Wikipedia itself states that Wikipedia isn't a source. Indeed anyone can write anything there, and anyone can edit it again. Although Wikipedia articles use to be somewhat reliable (but not always), they can't be trusted blindly and you have to rely on the references instead. "According to Wikipedia" means "according to an anonymous user who edited the article" (sometimes the mistakes are found months afterwards), so it should be "according to [Wikipedia references]" that you can verify.
        As for what "LindaB" tells, you can find loads of article by searching for "wikipedia medical" in Google News.
        Anyway the part quoted by "Unknown" is supported by a reference to the website... but it can be no longer verified, as the page has been removed and redirected to a different page. Moreover, that Planned Features section has been added just May 28th (see history of the page). Maybe it is correct anyway, but it can't be verified now.

      • Gel · 58 days ago

        Wikipedia is most definitely a valid and reliable source, not 100% but then again, nothing save for definitive scientific publications are. Those years of teachers trying to scare you into not using Wikipedia sure have stuck with you.

  4. Itoldem · 61 days ago

    Sabu

    • Anonymous · 61 days ago

      This. Right here. Add in a bit "National Security"...

  5. This all feels very weird to me. No official statement saying it'd been hacked...nothing anywhere from TrueCrypt on this apparent defacement & hacking.

    I don't trust any of it. Not on my main computer but I don't know what I could find!

    @fewwordswoman - self-taught programmer & infosec enthusiast!

  6. First of all, DON'T PANIC !
    This is way too fishy to be trusted, so why take advice from them at this particular moment ? Anyone who knows about "Warrant Canary" will get the point. This the most logical thing that has happened at this moment. Wait for it, this is one hell of a story, and the truth will come out soon enough.

    • Paul Ducklin · 61 days ago

      I thought a "warrant canary" was the indication by means of something that *didn't* happen (e.g. *not* updating some sort of tell-tale "there has been no warrant yet") that law enforcement were on your case. Or are you suggesting that by saying a bunch of true stuff (of course the software has some unfixed bugs :-) the developers are implicitly adding some kind of unspoken claim?

      I suggest that if they are, it isn't terribly clear what it is that they're trying to say, and thus it hasn't worked, given all the other inferences - such as an unhealable rift in the development team causing them to give up as anonymously as they can - that might be drawn.)

      The problem with your comment is that it's kind of contradictory. You're saying that this is so fishy you should ignore the advice to stop using the product, yet if it really *is* that fishy, why would you continue to trust the product? (Which in turn might suggest it's a cruel hack designed to destroy it one way or another.)

      • Anonymous · 61 days ago

        Read Wikipedia: "The intention is to allow the provider to inform users of the existence of a subpoena passively without disclosing to others that the government has sought or obtained access to information or records under a secret subpoena." passively :)

        • LindaB · 61 days ago

          See comment above about Wikipedia

          • Anonymous · 59 days ago

            But you never provided a source about your Wikipedia comment.

        • rebelfud · 60 days ago

          I thought the M$ Bitlocker recommendation was the canary.

      • BigAl · 59 days ago

        And what if BitLocker was the program that got cracked, and the TrueCrypt was the hoax? Causing panicked people to needlessly expose their data?

        Any way you look at this fish, it smells bad.

  7. The source code of version 7.1 is available at GitHub:

  8. asm-wolf · 61 days ago

    A little bit more info:

    I found another article that has some more details, including a tweet from Matthew Green who was part of the audit. He claims to have no idea what is up, suggesting to me that the website changes are not legitimate.

    • Paul Ducklin · 61 days ago

      My own reading of Matthew Green's tweet, where he says that he has no idea what's up, is that he has no idea what's up, so please don't expect an explanation from him.

      In other words, one of the questions Matthew Green is rather specifically telling us he can't answer is whether this is legitimate or not :-)

      I suggest that all you can infer from Matthew Green's tweet is that he can't think of any audit-related reason why this might have happened.

  9. NSA is watching · 61 days ago

    If it's truly compromised it would be unusual not to just patch it (rather than scuttle the project entirely) The wierdness of the page could be a hasty way of knowing an NSL had been served.

    If it turns out to be a hacked site another option could be psychological warfare to discredit True Crypt. The hacked page was clearly bizarre. Why go through the extensive effort to hack both the site AND signing keys then botch it with such an obvious red flag?

  10. Joe · 61 days ago

    Not cool. What cross-platform alternatives will people be using?

  11. Anonymous · 61 days ago

    You don't go through the trouble of auditing something just to suddenly give up on it if everything looks OK. It's not like Microsoft's ending support for XP was a surprise. My interpretation is, "We DID find something major, but we're not going to say that yet, because it would expose all of you out there who are still using TrueCrypt. So here's something to help migrate to something else before we admit it."

    • Paul Ducklin · 61 days ago

      Thing is (at least for Windows users) it doesn't "help you migrate to something else," it pretty much instructs you to migrate to Bitlocker.

      As a representative of a company that sells a biggish suite of encryption products (I'll spare you a clickable link - this seems a poor place for one :-), I'm comfortable with closed source security software, but I still find it surprising for TrueCrypt to go out expressing what seems like a similar level of comfort in proprietary crypto code...

  12. Doing investigative journalism the lazy way: ask your readers what's going on.

    • Paul Ducklin · 61 days ago

      Firstly, I am not a journalist.

      Secondly, I actually put a fair amount of effort into this article at short notice, if you don't mind. (Oh, and a big thanks thanks to the guys in SophosLabs who chipped in at similrly short notice to help validate - so far as anything can be validated in this case - some of the assertions I made.)

      Thirdly, I felt that publishing my overview as quickly as I reasonably could was purposeful, given the sheer weirdness of the situation.

      Fourthly, this really is a case where "asking around" has considerable value, as you can see from the numerous comments that have been offered already.

      Fifthly, I am actually interested in what our readers think and have to say.

      Sixthly, what more is there to "investigate" at this point?

      (I was going to be ironic, and reply simply by writing "Commenting the lazy way: accuse the author of laziness." But I felt that a fuller explanation was relevant.)

      • Well said Paul. Keep up the awesome job over there at Sophos. Looking forward to the next podcast.

      • son-of-an-editor! · 60 days ago

        Firstly, you are not a journalist. Says it all doesn't it? :)

    • Anonymous · 60 days ago

      LMFAO. That's the funniest thing I've read all day.

  13. Not everyone can use Bitlocker. It's not built into the Home version of Win 7, for instance, and is not available from MS as an additional installation. For this and other reasons, I smell a rat.

    • I have a similar quandary. I'm cross platform - having a number of both Ubuntu and Windows installations. Ugh. This is bad news indeed.

  14. Al-Kammy · 61 days ago

    the claim that "Windows 8/7/Vista and later offer integrated support for encrypted disks and virtual disk images" is only true of the fully fledged windows versions. Anyone with the Home/Basic/Starter editions (prob'ly a VAST majority of joe public) are left to 3rd party suppliers, of which TruCrypt being free would have been most peoples first port of call

    I hope this turns out to be untrue,

  15. Shawn · 61 days ago

    So, truecrypt would be willing to recommend an unsecure OS, such as Windows as a new encryption method? I seriously doubt that. And why would having OS based encryption methods even be a viable reason to jump ship on truecrypt that is cross platform compatible? Using truecrypt allows you to access data on any OS that truecrypt was available for. I would never trust my stuff to MS encryption, whereas MS would be more than happy to give officials backdoor access when requested. Not that I hide anything illegal, but it is MY personal data that no one has any need to access.

    • Gel · 58 days ago

      Not only has Windows improved security over the years, but it's not any more secure than any other OS, there are simply more script kiddies pawing at Windows than the others.

  16. James · 61 days ago

    "Not
    Secure
    As"

    Hidden message ?

  17. Not
    Secure
    As

    hidden message ?

  18. A Robertson · 61 days ago

    Okay, deep breath everyone! Firstly Truecrypt was perfectly acceptable less than a week ago and you didn't lose sleep then (maybe over some other large organisation hack but not Truecrypt).

    The reality is that ANY encryption is better than no encryption as it immediately stops casual browsing and copying of data by ordinary people.

    Now, before everyone does the headless chicken act in blind panic and starts to remove it left, right and centre ask yourself what the possibilities and outcomes are:

    1) It is compromised and the person maintaining it knows that. The audit was going to find out and the whole project would be scrapped.

    Well so far the audit found nothing. I'm inclined to think that time is a great tester and had the encryption been cracked then people would have their suspicions and report it. I haven't heard of this - have you?

    2) It really does work and the authorities can't crack it.

    So, how do you get rid of it? Well FUD for starters. Panic sets in and everyone mistrusts it and gets rid of it. As usual, remove any previous versions and change the latest version to decrypt only. Then recommend a Microsoft product for encryption? Hmmmmmm.

    I would say if you are using it then keep using it - it's highly improbable that it has been compromised. Additionally let's wait until the audit is complete. Time will tell.

  19. TryMyBest · 61 days ago

    The end of TC is really bad news! The circumstances unfortunately give rise to lots of speculation.

    The most convincing scenario to me is that administrative authorities identified the coders. Some kind of legal threat including the instruction to NOT speak about it. As far as I understand, the coders want to let us know that especially the creation of encrypted volumes might be potentially compromised. Maybe, they had to hand over signing keys, whatever...
    Therefore, they created a goodbye version still allowing to migrate data, simultaneously with an indirect but very obvious warning to better not use it from now on even if upcoming TCs might claim, “All issues fixed”.

    I don't know anything, but this is how it makes sense to me.

  20. Janus Knudsen · 61 days ago

    It should be fairly doable to create a piece of software as TrueCrypt, however I do like the "Encrypt system partition/ drive" and that is piece of functionality is ingeniously developed.

  21. Puzzle · 61 days ago

    Another possibility is that it was discovered that one person working on the code is in cahoots with the NSA, or at influenced/strong-armed by the NSA. As we don't know who the team is, or where they are located, this is possible.

    "Using TrueCrypt is not secure as it may contain unfixed security issues"

    s/not secure as/NSA:

    "Using TrueCrypt is NSA. It may contain unfixed security issues."

    We know that NSA realizes it's easier to attack the implementation as opposed to the crypto itself.

    I know I'm reaching a bit here, but this is certainly puzzling.

    • I don't think you're far off to be honest. This length of time (in the infosec world) is long enough for some credible statement from TrueCrypt to be made.

      I'm gonna call governmental/authority intervention as well. Although if that's the case, then there'll be more uproar. In my opinion, disregarding anything Snowden/NSA/etc. related, I don't see this kind of software as being a major threat to security.

      But then if it IS authority intervention, why direct us to BitLocker & change the most recent key to decrypt only, going to all the trouble to encourage people with pretty little screenshots to migrate to similar software?

      Curiouser & curiouser...

  22. anonymous · 61 days ago

    Keep in mind Truecrypt 7.1a is completely safe. NSA wants to force users to switch to the 7.2 version which is neither safe nor functional. Just keep using 7.1a, it is not compromised and NSA hates it exactly for that reason.

    • Paul Ducklin · 60 days ago

      Actually, 7.2 is only a *decryptor* intended to get you off TrueCrypt.

  23. TRSS · 60 days ago

    Its a Hoax. Yes the site has been hijacked, yes the private key stolen, yes 7.2 is fake and not official. No TC is not finished, No there was not a gag-order or of any kind. The persons who did this are known and will be on DoX at some point, after the site is back under the control of the rightful owners. Ignore the hijacked page. Do not decrypt data or switch to Bitlocker or alike, TC Version 7.1a is safe and 7.2 is not.

    Ignore it and continue as normal, ignore news sites that tell you this is anything other than a hoax.

    • Paul Ducklin · 60 days ago

      Well...if the site has been hijacked and the private key stolen, it's not really a *hoax* saying that there are security issues with TrueCrypt.

      That's the problem with a product that is created anonymously, like TrueCrypt. No-one can declare the truth of the matter with any real authority (or at least with any credibility) now.

      I think it really is Game Over, one way or another.

      • That's very likely - but I would like someone with some authority to tell us whether it was an existing issue (that we've been secretly living with for two years), if they just got fed up with development and 7.1a is just fine for the moment, or if some three letter people in suits forced their hand and we need to don our tin-foil hats. I suppose some answers will come as Mr Green finishes his audit - but that will only really answer some of the questions; the deviation of the Dev's MO is what bothers me the most though.

  24. ungars · 60 days ago

    "TrueCrypt - Free Open-Source On-The-Fly Disk Encryption" : the code source is in the hands of many people. This software will never die.

  25. Arch · 60 days ago

    Wayback - no cache for http://www.truecrypt.org
    Google - no cache for http://www.truecrypt.org
    Coral - no cache for http://www.truecrypt.org

    • At the risk of being accused of wearing a tin foil hat (which isn't so bad...they're nice & shiny after all)....

      ....that sounds to me like more evidence it's to do with some government or ~security~ intervention as opposed to hacking.

  26. Hearth · 60 days ago

    As far as I am concerned, I say keep using it (I will) until such time as the audit gives us a reason not to - it may prove everything is fine, and the Linux Foundation may fork it as indicated they are considering.

  27. When you study the case of Lavabit and hear what Ladar Levison has to say and Merril say, you will understand how the US government do to their inhabitants.
    There is mass surveillance and no privacy. When there are programs with privacy, they are trying to put it down or to get information which is the privacy.
    Merril fought several years to tell about that he got National Security Letter and how the US government or FBI works.
    When you hear about this you understand what happened to True Crypt and the reason that we have not gotten more information.

  28. Gel · 58 days ago

    I don't understand the need to remove TrueCrypt, not only would it be fallacious logic to assume that any other software is more secure, but of course nobody expected TC to be perfectly secure, it's still more secure than having nothing at all, so removing it will give no benefit at all.

  29. Anonymous · 58 days ago

    What is True Crypt and why do we need it?

  30. Reggie · 57 days ago

    Close down open-source. Form a company and fork out new brand with same software. Make it paid software and closed code. Done! Sell and be sold to security agencies.

  31. Would FVE software such as BitLocker, DDPE, DESLock require some form of authority backdoor to overcome the less savoury sort who use the software to cover up their tracks? Maybe they were forced offline by impending cases?

    • Paul Ducklin · 57 days ago

      In at least some jurisdictions, that issue is dealt with (to some extent) by laws requiring you to give up your password when asked, and penalising you if you don't. So it depends whether the penalty for not giving up your password is better or worse than giving it up.

      Similar sort of impasse was solved in many countries for enforcing drink driving laws, by making the penalty for refusing to submit to an alcohol breath test equal to or worse than breathing into the machine and coming up over the limit.

      • Deonast · 57 days ago

        I often wonder what happens if you say you've forgotten your password. Do you have to prove you've forgotten or do they have to prove you haven't, by some sort of vulcan mind meld.

        • Paul Ducklin · 57 days ago

          Don't know. I presume it then depends on whether the magistrate or the jury believes you...

          • Concerned Observer · 47 days ago

            This is why you don't memorise your password. You have it machine generated and placed on something like a Yubikey. Lose/destroy the Yubikey and there is no way, even under torture, that you could provide the password.

  32. JimLink · 21 days ago

    Folks, I think I've solved the mystery! The TrueCrypt devs were on flight MH-370!!

  33. Users:
    TrueCrypt has never been secure. I am a government IT contractor and I have personally seen TrueCrypt defeated within a matter of minutes using some something that I cannot comment on (this was three years ago)

    However, the PC's did not have the TPM on them; and I am not sure if that would have made entry into the device harder. The devices were not decrypted but there was a way to bypass the full disk encryption.

    Any thought out there?

    • JuicySluice · 13 days ago

      Any thought on something that you brought up but "cannot comment on"? Yeah, I have one or two, but they're NSFW.

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

About the author

Paul Ducklin is a passionate security proselytiser. (That's like an evangelist, but more so!) He lives and breathes computer security, and would be happy for you to do so, too. Paul won the inaugural AusCERT Director's Award for Individual Excellence in Computer Security in 2009. Follow him on Twitter: @duckblog