]> Hazelcast

neil.e.madden@gmail.com

Security Javascript Object Signing and Encryption Internet-Draft This document updates to deprecate the JWS algorithm "none" and the JWE algorithm "RSA1_5". These algorithms have known security weaknesses. It also updates the Review Instructions for Designated Experts to establish baseline security requirements that future algorithm registrations are expected to meet. About This Document The latest revision of this draft can be found at . Status information for this document may be found at . Discussion of this document takes place on the Javascript Object Signing and Encryption Working Group mailing list (), which is archived at . Subscribe at . Source for this draft and an issue tracker can be found at .

Introduction JSON Web Algorithms (JWA, ) introduced several standard algorithms for both JSON Web Signature (JWS) and JSON Web Encryption (JWE). Many of these algorithms have stood the test of time and are still in widespread use. However, some algorithms have proved to be difficult to implement correctly leading to exploitable vulnerabilities. This document deprecates two such algorithms:

• The JWS "none" algorithm, which indicates that no security is applied to the message at all.
• The JWE "RSA1_5" algorithm, which indicates RSA encryption with PKCS#1 version 1.5 padding.

Note that RSA signatures using PKCS#1 version 1.5 padding ("RS256", "RS384", and "RS512") are unchanged by this specification and can still be used. Additionally, this document also updates the Review Instructions for the JOSE Designated Experts, to establish baseline security requirements for future JOSE algorithm registrations. Only algorithms that are reasonably believed to satisfy these requirements are expected to be registered in the future.

The 'none' algorithm The "none" algorithm creates an Unsecured JWS, whose contents are completely unsecured as the name implies. Despite strong guidance in the original RFC around not accepting Unsecured JWS by default, many implementations have had serious bugs due to accepting this algorithm. In some cases, this has led to a complete loss of security as authenticity and integrity checking can be disabled by an adversary simply by changing the algorithm ("alg") header in the JWS. The website tracks public vulnerabilities due to implementations mistakenly accepting the "none" algorithm. At the time of writing it lists 17 reports, many of which have high impact ratings. The following is a partial list of issues known to have been caused by misuse of the "none" algorithm, with a Common Vulnerability Enumeration identifier, and a Common Vulnerability Scoring System score indicating the severity of the impact:

• CVE-2017-10862 - CVSS: 5.3 (Medium)
• CVE-2018-1000531 - CVSS: 7.5 (High)
• CVE-2020-15957 - CVSS: 7.5 (High)
• CVE-2021-22160 - CVSS: 9.8 (Critical)
• CVE-2021-29500 - CVSS: 7.5 (High)
• CVE-2021-29451 - CVSS: 9.1 (Critical)
• CVE-2021-29455 - CVSS: 7.5 (High)
• CVE-2021-32631 - CVSS: 6.5 (Medium)
• CVE-2022-23540 - CVSS: 7.6 (High)
• CVE-2023-29357 - CVSS: 9.8 (Critical)

Many other vulnerabilities have been reported without an accompanying CVE, which we do not list here. Although there are some historical use-cases for Unsecured JWS that are not security vulnerabilities, these are relatively few in number and can easily be satisfied by alternative means. For example, two of these are in OpenID Connect : (1) securing unsigned ID Tokens via transmission over TLS in Section 3.1.3.7 and (2) the use of unsigned request objects in Section 6.1. The small risk of breaking some of these use-cases is far outweighed by the improvement in security for the majority of JWS users who may be impacted by accidental acceptance of the "none" algorithm.

The 'RSA1_5' algorithm The "RSA1_5" algorithm implements RSA encryption using PKCS#1 version 1.5 padding . This padding mode has long been known to have security issues, since at least Bleichenbacher's attack in 1998. It was supported in JWE due to the wide deployment of this algorithm, especially in legacy hardware. However, more secure replacements such as OAEP or elliptic curve encryption algorithms are now widely available. NIST has disallowed the use of this encryption mode for federal use since the end of 2023 and a CFRG draft also deprecates this encryption mode for new protocols and deployments. This document therefore also deprecates this algorithm for JWE.

Guidance on deprecation Both of the algorithms listed above are deprecated for use in JOSE—the "none" algorithm for JWS, and "RSA1_5" for JWE. JOSE library developers SHOULD deprecate support for these algorithms. Application developers MUST disable support for these algorithms by default. Consistent with the existing requirement in that implementations "MUST NOT accept Unsecured JWSs by default", an application that has a specific need for one of these algorithms MAY enable it, but only for the specific objects or operations that require it and not at a global level. New specifications building on top of JOSE MUST NOT allow the use of either algorithm. The IANA algorithm registry distinguishes between algorithms that are "Deprecated" and those that are "Prohibited". The algorithms identified in this document are to be marked as Deprecated only. Existing specifications and applications that make use of these algorithms can continue to do so, but are encouraged to adopt alternatives in future updates.

Conventions and Definitions The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "NOT RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in BCP 14 when, and only when, they appear in all capitals, as shown here.

Security Considerations This entire document is concerned with security, since the security of JOSE implementations directly affects the security of systems that include them (see for example the long list of CVEs in ).

IANA Considerations

JOSE Algorithm Deprecations The following changes are to be made to the IANA JOSE Web Signature and Encryption Algorithms registry:

• For the entry with Algorithm Name "none", update the JOSE Implementation Requirements to "Deprecated".
• For the entry with Algorithm Name "RSA1_5", update the JOSE Implementation Requirements to "Deprecated".

Updated Review Instructions for Designated Experts The review instructions for the designated experts for the IANA "JSON Web Signature and Encryption Algorithms" registry in are updated to add the following review criteria. As with the existing criteria in , these criteria do not apply to algorithms being registered as Deprecated or Prohibited.

• For JWS signature and MAC algorithms (the "alg" parameter values used with JWS), only algorithms that are reasonably believed to meet the standard security goal of existential unforgeability under a chosen message attack (EUF-CMA) are to be approved. See textbooks such as (Section 13.1.1) for a definition of existential unforgeability.
• For JWE key management algorithms (the "alg" parameter values used with JWE), only algorithms for which the resulting JWE encryption process as a whole is reasonably believed to meet the standard security goal of indistinguishability under an adaptive chosen ciphertext attack (IND-CCA2) are to be approved. This goal applies to the entire JWE encryption process and not to the key management algorithm in isolation. See textbooks such as (Section 9.2.2 and Chapter 12).
• For JWE content encryption methods (the "enc" parameter values used with JWE), only algorithms that are reasonably believed to meet the standard security goal of authenticated encryption with associated data (AEAD) are to be approved. See , and textbooks such as (Section 9.1), for the definition of AEAD security.

References Normative References This specification registers cryptographic algorithms and identifiers to be used with the JSON Web Signature (JWS), JSON Web Encryption (JWE), and JSON Web Key (JWK) specifications. It defines several IANA registries for these identifiers. This document defines algorithms for Authenticated Encryption with Associated Data (AEAD), and defines a uniform interface and a registry for such algorithms. The interface and registry can be used as an application-independent set of cryptoalgorithm suites. This approach provides advantages in efficiency and security, and promotes the reuse of crypto implementations. [STANDARDS-TRACK] In many standards track documents several words are used to signify the requirements in the specification. These words are often capitalized. This document defines these words as they should be interpreted in IETF documents. This document specifies an Internet Best Current Practices for the Internet Community, and requests discussion and suggestions for improvements. RFC 2119 specifies common key words that may be used in protocol specifications. This document aims to reduce the ambiguity by clarifying that only UPPERCASE usage of the key words have the defined special meanings. Informative References This document provides recommendations for the implementation of public-key cryptography based on the RSA algorithm, covering cryptographic primitives, encryption schemes, signature schemes with appendix, and ASN.1 syntax for representing keys and for identifying the schemes. This document represents a republication of PKCS #1 v2.2 from RSA Laboratories' Public-Key Cryptography Standards (PKCS) series. By publishing this RFC, change control is transferred to the IETF. This document also obsoletes RFC 3447. Red Hat, Inc. This document specifies additions and amendments to RFC 8017. Specifically, it provides guidance to implementers of the standard to protect against side-channel attacks. It also deprecates the RSAES- PKCS-v1_5 encryption scheme, and provides an alternative depadding algorithm that protects against side-channel attacks raising from users of vulnerable APIs. The purpose of this specification is to increase security of RSA implementations. The document is a product of the Crypto Forum Research Group (CFRG). National Institute of Standards and Technology MITRE n.d. FIRST n.d. IANA

Acknowledgments The author would like to thank the following people for feedback and useful suggestions: Mike Ounsworth, Michael B. Jones, Yaron Sheffer, Brian Campbell, Aaron Parecki, Filip Skokan, Tim Bray, and John Mattsson.