The final control we consider in depth is secure e-mail. Think about how much you use e-mail and how much you rely on the accuracy of its contents. How would you react if you received a message from your instructor saying that because you had done so well in your course so far, you were excused from doing any further work in it? What if that message were a joke from a classmate? We rely on e-mail's confidentiality and integrity for sensitive and important communications, even though ordinary e-mail has almost no confidentiality or integrity. In this section we investigate how to add confidentiality and integrity protection to ordinary e-mail.
Security for E-Mail
E-mail is vital for today's commerce, as well a convenient medium for communications among ordinary users. But, as we noted earlier, e-mail is very public, exposed at every point from the sender's workstation to the recipient's screen. Just as you would not put sensitive or private thoughts on a postcard, you must also acknowledge that e-mail messages are exposed and available for others to read.
Sometimes we would like e-mail to be more secure. To define and implement a more secure form, we begin by examining the exposures of ordinary e-mail.
Threats to E-Mail
Consider threats to electronic mail:
- message interception (confidentiality)
- message interception (blocked delivery)
- message interception and subsequent replay
- message content modification
- message origin modification
- message content forgery by outsider
- message origin forgery by outsider
- message content forgery by recipient
- message origin forgery by recipient
- denial of message transmission
Confidentiality and content forgery are often handled by encryption. Encryption can also help in a defense against replay, although we would also have to use a protocol in which each message contains something unique that is encrypted. Symmetric encryption cannot protect against forgery by a recipient, since both sender and recipient share a common key; however, public key schemes can let a recipient decrypt but not encrypt. Because of lack of control over the middle points of a network, senders or receivers generally cannot protect against blocked delivery.
Requirements and Solutions
If we were to make a list of the requirements for secure e-mail, our wish list would include the following protections.
message confidentiality (the message is not exposed en route to the receiver)
message integrity (what the receiver sees is what was sent)
sender authenticity (the receiver is confident who the sender was)
nonrepudiation (the sender cannot deny having sent the message)
Not all of these qualities are needed for every message, but an ideal secure e-mail package would allow these capabilities to be invoked selectively.
The standard for encrypted e-mail was developed by the Internet Society, through its architecture board (IAB) and research (IRTF) and engineering (IETF) task forces. The encrypted e-mail protocols are documented as an Internet standard in documents 1421, 1422, 1423, and 1424 [LIN93, KEN93, BAL93, KAL93a]. This standard is actually the third refinement of the original specification.
One of the design goals for encrypted e-mail was allowing security-enhanced messages to travel as ordinary messages through the existing Internet e-mail system. This requirement ensures that the large existing e-mail network would not require change to accommodate security. Thus, all protection occurs within the body of a message.
Because the protection has several aspects, we begin our description of them by looking first at how to provide confidentiality enhancements. The sender chooses a (random) symmetric algorithm encryption key. Then, the sender encrypts a copy of the entire message to be transmitted, including FROM:, TO:, SUBJECT:, and DATE: headers. Next, the sender prepends plaintext headers. For key management, the sender encrypts the message key under the recipient's public key, and attaches that to the message as well. The process of creating an encrypted e-mail message is shown in Figure 7-44.
Figure 7-44 Overview of Encrypted E-Mail Processing.
Encryption can potentially yield any string as output. Many e-mail handlers expect that message traffic will not contain characters other than the normal printable characters. Network e-mail handlers use unprintable characters as control signals in the traffic stream. To avoid problems in transmission, encrypted e-mail converts the entire ciphertext message to printable characters. An example of an encrypted e-mail message is shown in Figure 7-45. Notice the three portions: an external (plaintext) header, a section by which the message encryption key can be transferred, and the encrypted message itself. (The encryption is shown with shading.)
Figure 7-45 Encrypted E-MailSecured Message.
The encrypted e-mail standard works most easily as just described, using both symmetric and asymmetric encryption. The standard is also defined for symmetric encryption only: To use symmetric encryption, the sender and receiver must have previously established a shared secret encryption key. The processing type ("Proc-Type") field tells what privacy enhancement services have been applied. In the data exchange key field ("DEK-Info"), the kind of key exchange (symmetric or asymmetric) is shown. The key exchange ("Key-Info") field contains the message encryption key, encrypted under this shared encryption key. The field also identifies the originator (sender) so that the receiver can determine which shared symmetric key was used. If the key exchange technique were to use asymmetric encryption, the key exchange field would contain the message encryption field, encrypted under the recipient's public key. Also included could be the sender's certificate (used for determining authenticity and for generating replies).
The encrypted e-mail standard is designed to support multiple encryption algorithms, using popular algorithms such as DES, triple DES, and AES for message confidentiality, and RSA and DiffieHellman for key exchange.
Other Security Features
In addition to confidentiality, we may want various forms of integrity for secure e-mail.
Encrypted e-mail messages always carry a digital signature, so the authenticity and nonrepudiability of the sender is assured. The integrity is also assured because of a hash function (called a message integrity check, or MIC) in the digital signature. Optionally, encrypted e-mail messages can be encrypted for confidentiality.
Notice in Figure 7-45 that the header inside the message (in the encrypted portion) differs from that outside. A sender's identity or the actual subject of a message can be concealed within the encrypted portion.
The encrypted e-mail processing can integrate with ordinary e-mail packages, so a person can send both enhanced and nonenhanced messages, as shown in Figure 7-46. If the sender decides to add enhancements, an extra bit of encrypted e-mail processing is invoked on the sender's end; the receiver must also remove the enhancements. But without enhancements, messages flow through the mail handlers as usual.
Figure 7-46 Encrypted E-Mail Processing in Message Transmission.
S/MIME (discussed later in this section) can accommodate the exchange of other than just text messages: support for voice, graphics, video, and other kinds of complex message parts.
Encryption for Secure E-Mail
The major problem with encrypted e-mail is key management. The certificate scheme described in Chapter 2 is excellent for exchanging keys and for associating an identity with a public encryption key. The difficulty with certificates is building the hierarchy. Many organizations have hierarchical structures. The encrypted e-mail dilemma is moving beyond the single organization to an interorganizational hierarchy. Precisely because of the problem of imposing a hierarchy on a nonhierarchical world, PGP was developed as a simpler form of encrypted e-mail.
Encrypted e-mail provides strong end-to-end security for electronic mail. Triple DES, AES, and RSA cryptography are quite strong, especially if RSA is used with a long bit key (1024 bits or more). The vulnerabilities remaining with encrypted e-mail come from the points not covered: the endpoints. An attacker with access could subvert a sender's or receiver's machine, modifying the code that does the privacy enhancements or arranging to leak a cryptographic key.
Example Secure E-Mail Systems
Encrypted e-mail programs are available from many sources. Several universities (including Cambridge University in England and the University of Michigan in the United States) and companies (BBN, RSA-DSI, and Trusted Information Systems) have developed either prototype or commercial versions of encrypted e-mail.
PGP stands for Pretty Good Privacy. It was invented by Phil Zimmerman in 1991. Originally a free package, it became a commercial product after being bought by Network Associates in 1996. A freeware version is still available. PGP is widely available, both in commercial versions and freeware, and it is heavily used by individuals exchanging private e-mail.
PGP addresses the key distribution problem with what is called a "ring of trust" or a user's "keyring." One user directly gives a public key to another, or the second user fetches the first's public key from a server. Some people include their PGP public keys at the bottom of e-mail messages. And one person can give a second person's key to a third (and a fourth, and so on). Thus, the key association problem becomes one of caveat emptor: "let the buyer beware." If I am reasonably confident that an e-mail message really comes from you and has not been tampered with, I will use your attached public key. If I trust you, I may also trust the keys you give me for other people. The model breaks down intellectually when you give me all the keys you received from people, who in turn gave you all the keys they got from still other people, who gave them all their keys, and so forth.
You sign each key you give me. The keys you give me may also have been signed by other people. I decide to trust the veracity of a key-and-identity combination, based on who signed the key.
PGP does not mandate a policy for establishing trust. Rather, each user is free to decide how much to trust each key received.
The PGP processing performs some or all of the following actions, depending on whether confidentiality, integrity, authenticity, or some combination of these is selected:
Create a random session key for a symmetric algorithm.
Encrypt the message, using the session key (for message confidentiality).
Encrypt the session key under the recipient's public key.
Generate a message digest or hash of the message; sign the hash by encrypting it with the sender's private key (for message integrity and authenticity).
Attach the encrypted session key to the encrypted message and digest.
Transmit the message to the recipient.
The recipient reverses these steps to retrieve and validate the message content.
An Internet standard governs how e-mail is sent and received. The general MIME specification defines the format and handling of e-mail attachments. S/MIME (Secure Multipurpose Internet Mail Extensions) is the Internet standard for secure e-mail attachments.
S/MIME is very much like PGP and its predecessors, PEM (Privacy-Enhanced Mail) and RIPEM [LIN93, KEN93, BAL93, KAL93a]. The Internet standards documents defining S/MIME (version 3) are described in [HOU99] and [RAM99]. S/MIME has been adopted in commercial e-mail packages, such as Eudora and Microsoft Outlook.
The principal difference between S/MIME and PGP is the method of key exchange. Basic PGP depends on each user's exchanging keys with all potential recipients and establishing a ring of trusted recipients; it also requires establishing a degree of trust in the authenticity of the keys for those recipients. S/MIME uses hierarchically validated certificates, usually represented in X.509 format, for key exchange. Thus, with S/MIME, the sender and recipient do not need to have exchanged keys in advance as long as they have a common certifier they both trust.
S/MIME works with a variety of cryptographic algorithms, such as DES, AES, and RC2 for symmetric encryption.
S/MIME performs security transformations very similar to those for PGP. PGP was originally designed for plaintext messages, but S/MIME handles (secures) all sorts of attachments, such as data files (for example, spreadsheets, graphics, presentations, movies, and sound). Because it is integrated into many commercial e-mail packages, S/MIME is likely to dominate the secure e-mail market.