TrueCrypt

» see truecrypt.org as it lived, on the wayback machine

~ the eulogy ~

i used truecrypt around 2010 on a thinkpad in turkey, the way most engineers used it: full-disk encryption on the work laptop and an encrypted container on the side for everything else. the tool was free, ran on the major operating systems of the era, and worked the way crypto tools are supposed to work, which means you configure it once and forget about it until you reboot and it asks for your password.

the project started in february 2004 as a fork of an older program called e4m. it was maintained by a group calling itself the truecrypt foundation, and the actual humans behind it stayed anonymous for the entire ten-year run. the foundation accepted donations through a paypal account and published the source. the binaries were signed. nobody outside a small circle could put names to faces. trademark filings later identified two people named ondrej tesarik and david tesarik, but even that was a thin lead.

the tool was a workhorse. you picked an encryption algorithm out of a list that included aes and serpent. you set a passphrase. you got back either a virtual disk that mounted as a regular drive once unlocked, or a full-disk encrypted boot setup that asked for the passphrase before windows started. the design was clean enough that the fbi reportedly failed to break it on at least two seized machines without the passphrase. for a free tool from anonymous authors that was an unusual track record.

~ the dossier ~

~ the warning page ~

on may 28, 2014, the truecrypt.org domain started serving a single static page. the headline read "WARNING: Using TrueCrypt is not secure as it may contain unfixed security issues." the body said development had ended after microsoft ended support for windows xp. it recommended users migrate their data to bitlocker on windows, or to other tools on linux and macos. the final version 7.2 it shipped alongside the page could only decrypt existing volumes, not create new ones. the announcement was unsigned. it appeared all at once on a sunday.

the security community read this two ways. the surface reading was that the anonymous maintainers were tired and quitting. the deeper reading was that something had happened that they could not openly describe. the page recommended bitlocker, which is a closed-source microsoft tool. that recommendation made very little sense coming from people who had spent ten years writing free open-source full-disk encryption. the warrant canary theory took hold quickly: the developers had received a national-security letter or equivalent legal pressure that they could not legally describe, and had communicated through what they would and would not say.

~ the audits ~

a community-funded audit of truecrypt was already in progress when the shutdown happened. the open crypto audit project had crowd-funded $70,000 and engaged ncc group's cryptography services to do a formal review. the audit ran in two phases. phase one covered the bootloader and windows kernel driver and finished in april 2014, before the shutdown. it found no backdoors and four issues of varying severity. phase two covered the cryptographic implementation and finished in april 2015. it also found no backdoors and no severe design flaws.

the audits did not answer why the project ended the way it did. they only answered the technical question. by the time phase two completed, two open-source forks were already running. veracrypt picked up the codebase and is the most active descendant in 2026. ciphershed took a different path and faded quickly. the actual truecrypt repository on truecrypt.org stayed frozen as a memorial to the warning page.

~ what it left ~

the unanswered question is the legacy. truecrypt is the most-discussed case of an anonymous open-source security project that disappeared in a way nobody could explain, and the case files are still open in 2026. the technical work survived in veracrypt, which is what i would install today if i needed full-disk encryption on a machine that does not come with a clean built-in option. the political question, whether a free encryption tool can survive contact with state-actor pressure, has not gotten a cleaner answer since.

@}-->--

              )
             (_)
            .-'-.
            |   |
            |   |
            |   |
            |   |
          __|   |__   .-.
       .-'  |   |  `-:   :
      :     `---'     :-'
       `-._       _.-'
           '""""""

« previous  ·  random  ·  next »