Even though we're doing our best to offer you good tools to protect your privacy while using a computer, there is no magic or perfect solution to such a complex problem. Understanding well the limits of such tools is a crucial step in, first, deciding whether Tails is the right tool for you, and second, helping you making a good use of it.
- Tor exit nodes can eavesdrop on communications
- Tails makes it clear that you are using Tor and probably Tails
- Man-in-the-middle attacks
- Confirmation attacks
- Tails doesn't encrypt your documents by default
- Tails doesn't clear the metadata of your documents for you and doesn't encrypt the Subject: and other headers of your encrypted e-mail messages
- Tor doesn't protect you from a global adversary
- Tails doesn't magically separate your different contextual identities
- Tails doesn't make your crappy passwords stronger
- Tails is a work in progress
Tor is about hiding your location, not about encrypting your communication.
Instead of taking a direct route from source to destination, communications using the Tor network take a random pathway through several Tor relays that cover your tracks. So no observer at any single point can tell where the data came from or where it's going.
The last relay on this circuit, called the exit node, is the one that establishes the actual connection to the destination server. As Tor does not, and by design cannot, encrypt the traffic between an exit node and the destination server, any exit node is in a position to capture any traffic passing through it. See Tor FAQ: Can exit nodes eavesdrop on communications?.
For example, in 2007, a security researcher intercepted thousands of private e-mail messages sent by foreign embassies and human rights groups around the world by spying on the connections coming out of an exit node he was running. See Wired: Rogue Nodes Turn Tor Anonymizer Into Eavesdropper's Paradise.
To protect yourself from such attacks you should use end-to-end encryption.
Tails includes many tools to help you using strong encryption while browsing, sending email or chatting, as presented on our about page.
Your Internet Service Provider (ISP) or your local network administrator can see that you're connecting to a Tor relay, and not a normal web server for example. Using Tor bridges in certain conditions can help you hide the fact that you are using Tor.
The destination server that you are contacting through Tor can know whether your communication comes out from a Tor exit node by consulting the publicly available list of exit nodes that might contact it. For example using the Tor Bulk Exit List tool of the Tor Project.
So using Tails doesn't make you look like any random Internet user. The anonymity provided by Tor and Tails works by trying to make all of their users look the same so it's not possible to identify who is who amongst them.
A man-in-the-middle attack (MitM) is a form of active eavesdropping in which the attacker makes independent connections with the victims and relays messages between them, making them believe that they are talking directly to each other over a private connection, when in fact the entire conversation is controlled by the attacker.
While using Tor, man-in-the-middle attacks can still happen between the exit node and the destination server. The exit node itself can also act as a man-in-the-middle. For an example of such an attack see MW-Blog: TOR exit-node doing MITM attacks.
Again, to protect yourself from such attacks you should use end-to-end encryption and while doing so taking extra care at verifying the server authenticity.
Usually, this is automatically done throught SSL certificates checked by your browser against a given set of recognized certificate authorities). If you get a security exception message such as this one you might be victim of a man-in-the-middle attack and should not bypass it unless you have another trusted way of checking the certificate's fingerprint with the people running the service.
But on top of that the certificate authorities model of trust on Internet is susceptible to various methods of compromise.
For example, on March 15, 2011, Comodo, one of the major SSL certificates company, reported that a user account with an affiliate registration authority had been compromised. It was then used to create a new user account that issued nine certificate signing requests for seven domains: mail.google.com, login.live.com, www.google.com, login.yahoo.com (three certificates), login.skype.com, addons.mozilla.org, and global trustee. See Comodo: The Recent RA Compromise.
Later in 2011, DigiNotar, a Dutch SSL certificate company, incorrectly issued certificates to a malicious party or parties. Later on, it came to light that they were apparently compromised months before or perhaps even in May of 2009 if not earlier. Rogue certificates were issued for domains such as google.com, mozilla.org, torproject.org, login.yahoo.com and many more. See, The Tor Project: The DigiNotar Debacle, and what you should do about it.
This still leaves open the possibility of a man-in-the-middle attack even when your browser is trusting an HTTPS connection.
On one hand, by providing anonymity, Tor makes it more difficult to perform a man-in-the-middle attack targeted at one specific person with the blessing of a rogue SSL certificate. But on the other end, Tor makes it easier for people or organizations running exit nodes to perform large scale MitM attempts, or attacks targeted at a specific server, and especially those among its users who happen to use Tor.
The Tor design doesn't try to protect against an attacker who can see or measure both traffic going into the Tor network and also traffic coming out of the Tor network. That's because if you can see both flows, some simple statistics let you decide whether they match up.
That could also be the case if your ISP (or your local network administrator) and the ISP of the destination server (or the destination server itself) cooperate to attack you.
Tor tries to protect against traffic analysis, where an attacker tries to learn whom to investigate, but Tor can't protect against traffic confirmation (also known as end-to-end correlation), where an attacker tries to confirm an hypothesis by monitoring the right locations in the network and then doing the math.
The documents that you might save on storage devices will not be encrypted by default, except in the encrypted persistent volume. But Tails provides you with tools to encrypt your documents, such as GnuPG, or encrypt your storage device, such as LUKS. It is likely that the files you may create will keep tracks that they were created using Tails.
If you need to access the local hard-disks of the computer you are using, be conscious that you might then leave trace of your activities with Tails on it.
Tails doesn't clear the metadata of your documents for you and doesn't encrypt the Subject: and other headers of your encrypted e-mail messages
Numerous files format store hidden data or metadata inside of the files. Text processors or PDF files could store the name of the author, the date and time of creation of the file, and sometimes even parts of the editing history of the file… those hidden data depend on the file format and the software used. Please note also, that the Subject: as well as the rest of the header lines of your OpenPGP encrypted e-mail messages are not encrypted. This is not a bug of Tails or the OpenPGP protocol; it's for backwards compatibility with the original SMTP protocol. Unfortunately no RFC standard exists yet for Subject encryption.
Images file formats, like TIFF of JPEG, probably take the prize in this field. Those files, created by digital cameras or mobile phones, contain a metadata format called EXIF which can include the date, time and sometimes the GPS coordinates of the picture, the brand and serial number of the device which took it as well as a thumbnail of the original image. Image processing software tend to keep those data intact. Internet is full of cropped or blurred images for which the EXIF thumbnail still contains the full original picture.
Tails doesn't clear the metadata of your files for you. Yet. Still it's in Tails' design goal to help you do that. For example, Tails already comes with the Metadata anonymisation toolkit.
A global passive adversary would be a person or an entity able to monitor at the same time the traffic between all the computers in a network. By studying, for example, the timing and volume patterns of the different communications across the network, it would be statistically possible to identify Tor circuits and thus matching Tor users and destination servers.
It is part of Tor's initial trade-off not to address such a threat in order to create a low-latency communication service usable for web browsing, Internet chat or SSH connections.
For more expert information see Tor Project: The Second-Generation Onion Router, part 3. Design goals and assumptions.
It is usually not advisable to use the same Tails session to perform two tasks or endorse two contextual identities that you really want to keep separate from another. For example hiding your location to check your email and publishing anonymously a document.
First, because Tor tends to reuse the same circuits, for example amongst a same browsing session. Since the exit node of a circuit knows both the destination server (and possibly the content of the communication if not encrypted) and the address of the previous relay it received the communication from, it makes it easier to correlate the several browsing requests as part of a same circuit and possibly made by a same user. If you are facing a global adversary as described above, it might then also be in position to do this correlation.
Second, in case of a security hole or a misuse in using Tails or one of its application, information about your session could be leaked. That could reveal that the same person was behind the various actions made during the session.
The solution to both threats is to shutdown and restart Tails every time you're using a new identity, if you really want to isolate them better.
Tor allows you to be anonymous online; Tails allows you to leave no trace on the computer you're using. But again, neither of both are magic spells for computer security.
If you use weak passwords, they can be guessed by brute-force attacks with or without Tails in the same way. To know if your passwords are weak and learn good practices to create better password, you can read Wikipedia: Weak Passwords.
Tails, as well as all the software it includes, are on continuous development and might contain programming errors or security holes. Stay tuned to Tails development.