Encryption Key Derivation in the Cryptographic Message Syntax (CMS) using HKDF with SHA-256

Status of This Memo

This Internet-Draft is submitted in full conformance with the provisions of BCP 78 and BCP 79.¶

Internet-Drafts are working documents of the Internet Engineering Task Force (IETF). Note that other groups may also distribute working documents as Internet-Drafts. The list of current Internet-Drafts is at https://datatracker.ietf.org/drafts/current/.¶

Internet-Drafts are draft documents valid for a maximum of six months and may be updated, replaced, or obsoleted by other documents at any time. It is inappropriate to use Internet-Drafts as reference material or to cite them other than as "work in progress."¶

This Internet-Draft will expire on 2 June 2024.¶

Copyright Notice

Copyright (c) 2023 IETF Trust and the persons identified as the document authors. All rights reserved.¶

This document is subject to BCP 78 and the IETF Trust's Legal Provisions Relating to IETF Documents (https://trustee.ietf.org/license-info) in effect on the date of publication of this document. Please review these documents carefully, as they describe your rights and restrictions with respect to this document. Code Components extracted from this document must include Revised BSD License text as described in Section 4.e of the Trust Legal Provisions and are provided without warranty as described in the Revised BSD License.¶

Table of Contents

1. Introduction

This document specifies the derivation of the content-encryption key for the Cryptographic Message Syntax (CMS) enveloped-data content type [RFC5652], the content-encryption key for the CMS encrypted-data content type [RFC5652], or the content-authenticated-encryption key for the authenticated-enveloped-data content type [RFC5083].¶

The use of this mechanism provides protection against where the attacker manipulates the content-encryption algorithm identifier or the content-authenticated-encryption algorithm identifier. Johannes Roth and Falko Strenzke presented such an attack at IETF 118 [RS2023], where:¶

1. The attacker intercepts a CMS Authenticated-Enveloped-Data content [RFC5083] that uses either AES-CCM or AES-GCM [RFC5084].¶

2. The attacker turns the intercepted content into a "garbage" CMS Enveloped-Data content Section 6 of [RFC5652] that is composed of AES-CBC guess blocks.¶

3. The attacker sends the "garbage" message to the victim, and the victim reveals the result of the decryption with the attacker.¶

4. If any of the transformed plaintext blocks match the guess for that block, then the attacker learns the plaintext for that block.¶

With highly structured messages, one block can reveal the only sensitive part of the original message.¶

This attack is thwarted if the encryption key depends upon the delivery of the unmodified algorithm identifier.¶

The mitigation for this attack has three parts:¶

• Potential recipients include the object identifier for the use-cms-cek-hkdf-sha256 attribute (with no parameters) in S/MIME Capabilities to indicate support for this mitigation.¶

• As a flag to the recipient that this mitigation is being used, carry the use-cms-cek-hkdf-sha256 attribute in the unprotected attributes.¶

• Perform encryption with a derived content-encryption key or content-authenticated-encryption key:¶

      CEK' = HKDF(CEK, AlgorithmIdentifier)

2. Use of of HKDF with SHA-256 to Derive Encryption Keys

The mitigation uses the HMAC-based Extract-and-Expand Key Derivation Function (HKDF) [RFC5869] to derive output keying materiam (OKM) from input key material (IKM). HKDF is used with the SHA-256 hash function [FIPS180]. The derivation includes the DER-encoded AlgoritmIdentifier as the optional info input value. If an attacker were to change the originator-provided AlgoritmIdentifier, then the recipient will derive a different content-encryption key or content-authenticated-encryption key.¶

The CMS_CEK_HKDF_SHA256 function uses the HKDF-Extract and HKDF-Expand functions to derive the OKM from the IKM:¶

Inputs:
   IKM      input keying material
   info     DER-encoded AlgoritmIdentifier

Output:
   OKM      output keying material (same size as IKM)

The output OKM is calculated as follows:

   OKM_SIZE = len(IKM)
   IF OKM_SIZE > 8160 THEN raise error

   salt = "The Cryptographic Message Syntax"
   PRK = HKDF-Extract(salt, IKM)

   OKM = HKDF-Expand(PRK, info, OKM_SIZE)

3. The use-cms-cek-hkdf-sha256 Attribute

The use-cms-cek-hkdf-sha256 attribute specifies that the CMS_CEK_HKDF_SHA256 function defined in Section 2 to derive the content-encryption key or the content-authenticated-encryption key.¶

The use-cms-cek-hkdf-sha256 attribute MUST be a unprotected attribute or an unauthenticated attribute; it MUST NOT be an signed attribute, authenticated attribute, or protected attribute.¶

The following object identifier identifies the use-cms-cek-hkdf-sha256 attribute:¶

   id-aa-use-cms-cek-hkdf-sha256 OBJECT IDENTIFIER ::= { TBD1 }

The use-cms-cek-hkdf-sha256 attribute values have ASN.1 type of NULL.¶

Using the conventions from [RFC5911], the use-cms-cek-hkdf-sha256 attribute is defined as:¶

  aa-useCMSCEKHKDFSHA256 ATTRIBUTE ::=
      { TYPE NULL IDENTIFIED BY id-aa-use-cms-cek-hkdf-sha256 }

4. SMIMECapabilities Attribute Conventions

The SMIMECapabilities Attribute is defined in Section 2.5.2 of [RFC8551]. An S/MIME client announces the set of cryptographic functions it supports using the SMIMECapabilities attribute.¶

If an S/MIME client is supports the mechanism in this document, the id-use-cms-cek-hkdf-sha256 OID SHOULD be included in the set of cryptographic functions. The parameter with this encoding MUST be absent.¶

The encoding for id-use-cms-cek-hkdf-sha256, in hexadecimal, is:¶

   30 TBD

5. Use of of HKDF with SHA-256 with CMS

This section describe the originator and recipient processing to implement this mitigation for each of the CMS encrypting content types.¶

4.  The content is encrypted with the content-encryption key.
    Content encryption may require that the content be padded to a
    multiple of some block size; see Section 6.3 of [RFC5652].

CEK' = CMS_CEK_HKDF_SHA256(CEK, ContentEncryptionAlgorithmIdentifier)

CEK' = CMS_CEK_HKDF_SHA256(CEK, ContentEncryptionAlgorithmIdentifier)

5.  The attributes collected in step 4 are authenticated and the CMS
    content is authenticated and encrypted with the content-
    authenticated-encryption key.  If the authenticated encryption
    algorithm requires either the additional authenticated data (AAD)
    or the content to be padded to a multiple of some block size,
    then the padding is added as described in Section 6.3 of
    [RFC5652].

CEK' = CMS_CEK_HKDF_SHA256(CEK, ContentEncryptionAlgorithmIdentifier)

6. Security Considerations

This mitigation always uses HKDF with SHA-256. One KDF algorithm was selected to avoid the need for negotiation. In the future, if a weakness is found in the KDF algorithm, a new attribute will need to be assigned for use with an alternative KDF algorithm.¶

If the attacker removes the use-cms-cek-hkdf-sha256 attribute from the unprotected attributes or unauthenticated attributes, then the recipient will derive a different CEK', which will not assist the attacker in recovering the plaintext content.¶

If the attacker changes the ContentEncryptionAlgorithmIdentifier, then the recipient will derive a different CEK', which will not assist the attacker in recovering the plaintext content. The AlgorithmIdentifier object identifier is sufficient to mitigate the attack described in [RS2023], but this mitigation includes the object identifier and the parameters to protect against some yet-to-be-discovered attack that only manipulates the parameters.¶

Implementations MUST protect the content-encryption keys and content-authenticated-encryption keys, this includes the CEK and CEK'. Compromise of a content-encryption key may result in disclosure of the associated encrypted content. Compromise of a content-authenticated-encryption key may result in disclosure of the associated encrypted content or allow modification of the authenticated content and the additional authenticated data (AAD).¶

Implementations MUST randomly generate content-encryption keys and content-authenticated-encryption keys. Using an inadequate pseudo-random number generator (PRNG) to generate cryptographic keys can result in little or no security. An attacker may find it much easier to reproduce the PRNG environment that produced the keys, and then searching the resulting small set of possibilities, rather than brute force searching the whole key space. The generation of quality random numbers is difficult. [RFC4086] offers important guidance on this topic.¶

7. Privacy Considerations

If the message-digest attribute is included in the AuthAttributes, then the attribute value will contain the unencrypted one-way hash value of the plaintext of the content. Disclosure of this hash value enables content tracking, and it can be used to determine if the plaintext matches one or more candidate contents. For these reasons, the AuthAttributes SHOULD NOT contain the message-digest attribute.¶

8. Operations Considerations

CMS is often used to provide encryption in messaging environments, where various forms of unsolicited messages (such as spam and phishing) represent a significant volume of unwanted traffic. Mitigation strategies for unwanted message traffic involve analysis of message plaintext. When recipients accept unsolicited encrypted messages, they become even more vulnerable to unwanted traffic since many mitigation strategies will be unable to access the message plaintext. Therefore, software that receives messages that have been encrypted using CMS ought to provide alternate mechanisms to handle the unwanted message traffic. One approach that does not require disclosure of keying material to a server is to reject or discard encrypted messages unless they purport to come from a member of a previously approve originator list.¶

9. IANA Considerations

For the ASN.1 Module in the Appendix A of this document, IANA is requested to assign an object identifier (OID) for the module identifier (TBD0) with a Description of "id-mod-CMS-CEK-HKDF-SHA256-2023". The OID for the module should be allocated in the "SMI Security for S/MIME Module Identifier" registry (1.2.840.113549.1.9.16.0).¶

For the use-cms-cek-hkdf-sha256 attribute definition in Section 3 of this document, IANA is requested to assign an object identifier (OID) (TBD1) with a Description of "id-aa-use-cms-cek-hkdf-sha256". The OID for the module should be allocated in the "SMI Security for
S/MIME Attributes" registry (1.2.840.113549.1.9.16.2).¶

10. Acknowledgements

Thanks to Mike Ounsworth for his careful review and constructive comments.¶

Appendix A. ASN.1 Module

This ASN.1 Module builds upon the conventions established in [RFC5911] and [RFC5912].¶

<CODE STARTS>

CMS-CEK-HKDF-SHA256-Module-2023
  { iso(1) member-body(2) us(840) rsadsi(113549) pkcs(1) pkcs9(9)
    id-smime(16) id-mod(0) id-mod-CMS-CEK-HKDF-SHA256-2023(TBD0) }

DEFINITIONS IMPLICIT TAGS ::= BEGIN

EXPORTS ALL;

IMPORTS
  ATTRIBUTE
  FROM PKIX-CommonTypes-2009   -- in [FRC5912]
    { iso(1) identified-organization(3) dod(6) internet(1)
      security(5) mechanisms(5) pkix(7) id-mod(0)
      id-mod-pkixCommon-02(57) }

  SMIME-CAPS
  FROM AlgorithmInformation-2009 -- in [FRC5911]
    { iso(1) identified-organization(3) dod(6) internet(1)
      security(5) mechanisms(5) pkix(7) id-mod(0)
      id-mod-algorithmInformation-02(58) } ;


--
-- use-CMS-CEK-HKDF-SHA256 Attribute
--

UnprotectedAttrs ATTRIBUTE ::= { aa-useCMSCEKHKDFSHA256, ... }

id-aa-use-cms-cek-hkdf-sha256 OBJECT IDENTIFIER ::= { TBD1 }

aa-useCMSCEKHKDFSHA256 ATTRIBUTE ::=
    { TYPE NULL IDENTIFIED BY id-aa-use-cms-cek-hkdf-sha256 }

--
-- S/MIIME Capability for use-CMS-CEK-HKDF-SHA256
--

SMimeCaps SMIME-CAPS ::= { cap-useCMSCEKHKDFSHA256, ... }

cap-useCMSCEKHKDFSHA256 SMIME-CAPS ::=
    { -- No value -- IDENTIFIED BY id-aa-use-cms-cek-hkdf-sha256 }

END

<CODE ENDS>

Appendix B. CMS_CEK_HKDF_SHA256 Function Examples

This appendix provides two test vectores for the CMS_CEK_HKDF_SHA256 function.¶

IKM = c702e7d0a9e064b09ba55245fb733cf3

The AES-128-CGM AlgorithmIdentifier:
 algorithm=2.16.840.1.101.3.4.1.6
 parameters=GCMParameters:
  aes-nonce=0x5c79058ba2f43447639d29e2
  aes-ICVlen is ommited; it indicates the DEFAULT of 12

DER-encoded AlgorithmIdentifier:
  301b0609608648016503040106300e040c5c79058ba2f43447639d29e2

OKM = 2124ffb29fac4e0fbbc7d5d87492bff3

IKM = c702e7d0a9e064b09ba55245fb733cf3

The AES-128-CBC AlgorithmIdentifier:
 algorithm=2.16.840.1.101.3.4.1.2
 parameters=AES-IV=0x651f722ffd512c52fe072e507d72b377

DER-encoded AlgorithmIdentifier:
  301d06096086480165030401020410651f722ffd512c52fe072e507d72b377

OKM = 9cd102c52f1e19ece8729b35bfeceb50

Author's Address