The Messaging Layer Security (MLS) ProtocolCiscorlb@ipv.sxInriabenjamin.beurdouche@inria.frFacebookjmillican@fb.comGoogleemadomara@google.comUniversity of Oxfordme@katriel.co.ukWireraphael@wire.com
Security
Internet-DraftMessaging applications are increasingly making use of end-to-end
security mechanisms to ensure that messages are only accessible to
the communicating endpoints, and not to any servers involved in delivering
messages. Establishing keys to provide such protections is
challenging for group chat settings, in which more than two
clients need to agree on a key but may not be online at the same
time. In this document, we specify a key establishment
protocol that provides efficient asynchronous group key establishment
with forward secrecy and post-compromise security for groups
in size ranging from two to thousands.DISCLAIMER: This is a work-in-progress draft of MLS and has not yet
seen significant security analysis. It should not be used as a basis
for building production systems.RFC EDITOR: PLEASE REMOVE THE FOLLOWING PARAGRAPH The source for
this draft is maintained in GitHub. Suggested changes should be
submitted as pull requests at https://github.com/mlswg/mls-protocol.
Instructions are on that page as well. Editorial changes can be
managed in GitHub, but any substantive change should be discussed on
the MLS mailing list.A group of users who want to send each other encrypted messages needs
a way to derive shared symmetric encryption keys. For two parties,
this problem has been studied thoroughly, with the Double Ratchet
emerging as a common solution .
Channels implementing the Double Ratchet enjoy fine-grained forward secrecy
as well as post-compromise security, but are nonetheless efficient
enough for heavy use over low-bandwidth networks.For a group of size greater than two, a common strategy is to
unilaterally broadcast symmetric "sender" keys over existing shared
symmetric channels, and then for each member to send messages to the
group encrypted with their own sender key. Unfortunately, while this
improves efficiency over pairwise broadcast of individual messages and
provides forward secrecy (with the addition of a hash ratchet),
it is difficult to achieve post-compromise security with
sender keys. An adversary who learns a sender key can often indefinitely and
passively eavesdrop on that member's messages. Generating and
distributing a new sender key provides a form of post-compromise
security with regard to that sender. However, it requires
computation and communications resources that scale linearly with
the size of the group.In this document, we describe a protocol based on tree structures
that enable asynchronous group keying with forward secrecy and
post-compromise security. Based on earlier work on "asynchronous
ratcheting trees" , the protocol presented here uses an
asynchronous key-encapsulation mechanism for tree structures.
This mechanism allows the members of the group to derive and update
shared keys with costs that scale as the log of the group size.RFC EDITOR PLEASE DELETE THIS SECTION.draft-10Allow new members to join via an external Commit (*)Enable proposals to be sent inline in a Commit (*)Re-enable constant-time Add (*)Change expiration extension to lifetime extension (*)Make the tree in the Welcome optional (*)PSK injection, re-init, sub-group branching (*)Require the initial init_secret to be a random value (*)Remove explicit sender data nonce (*)Do not encrypt to joiners in UpdatePath generation (*)Move MLSPlaintext signature under the confirmation tag (*)Explicitly authenticate group membership with MLSPLaintext (*)Clarify X509Credential structure (*)Remove uneeded interim transcript hash from GroupInfo (*)IANA considerationsDerive an authentication secretUse Extract/Expand from HPKE KDFClarify that application messages MUST be encrypteddraft-09Remove blanking of nodes on Add (*)Change epoch numbers to uint64 (*)Add PSK inputs (*)Add key schedule exporter (*)Sign the updated direct path on Commit, using "parent hashes" and one
signature per leaf (*)Use structured types for external senders (*)Redesign Welcome to include confirmation and use derived keys (*)Remove ignored proposals (*)Always include an Update with a Commit (*)Add per-message entropy to guard against nonce reuse (*)Use the same hash ratchet construct for both application and handshake keys (*)Add more ciphersuitesUse HKDF to derive key pairs (*)Mandate expiration of ClientInitKeys (*)Add extensions to GroupContext and flesh out the extensibility story (*)Rename ClientInitKey to KeyPackagedraft-08Change ClientInitKeys so that they only refer to one ciphersuite (*)Decompose group operations into Proposals and Commits (*)Enable Add and Remove proposals from outside the group (*)Replace Init messages with multi-recipient Welcome message (*)Add extensions to ClientInitKeys for expiration and downgrade resistance (*)Allow multiple Proposals and a single Commit in one MLSPlaintext (*)draft-07Initial version of the Tree based Application Key Schedule (*)Initial definition of the Init message for group creation (*)Fix issue with the transcript used for newcomers (*)Clarifications on message framing and HPKE contexts (*)draft-06Reorder blanking and update in the Remove operation (*)Rename the GroupState structure to GroupContext (*)Rename UserInitKey to ClientInitKeyResolve the circular dependency that draft-05 introduced in the
confirmation MAC calculation (*)Cover the entire MLSPlaintext in the transcript hash (*)draft-05Common framing for handshake and application messages (*)Handshake message encryption (*)Convert from literal state to a commitment via the "tree hash" (*)Add credentials to the tree and remove the "roster" concept (*)Remove the secret field from tree node valuesdraft-04Updating the language to be similar to the Architecture documentECIES is now renamed in favor of HPKE (*)Using a KDF instead of a Hash in TreeKEM (*)draft-03Added ciphersuites and signature schemes (*)Re-ordered fields in UserInitKey to make parsing easier (*)Fixed inconsistencies between Welcome and GroupState (*)Added encryption of the Welcome message (*)draft-02Removed ART (*)Allowed partial trees to avoid double-joins (*)Added explicit key confirmation (*)draft-01Initial description of the Message Protection mechanism. (*)Initial specification proposal for the Application Key Schedule
using the per-participant chaining of the Application Secret design. (*)Initial specification proposal for an encryption mechanism to protect
Application Messages using an AEAD scheme. (*)Initial specification proposal for an authentication mechanism
of Application Messages using signatures. (*)Initial specification proposal for a padding mechanism to improving
protection of Application Messages against traffic analysis. (*)Inversion of the Group Init Add and Application Secret derivations
in the Handshake Key Schedule to be ease chaining in case we switch
design. (*)Removal of the UserAdd construct and split of GroupAdd into Add
and Welcome messages (*)Initial proposal for authenticating handshake messages by signing
over group state and including group state in the key schedule (*)Added an appendix with example code for tree mathChanged the ECIES mechanism used by TreeKEM so that it uses nonces
generated from the shared secretdraft-00Initial adoption of draft-barnes-mls-protocol-01 as a WG item.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.
An agent that uses this protocol to establish shared cryptographic
state with other clients. A client is defined by the
cryptographic keys it holds.
A collection of clients with shared cryptographic state.
A client that is included in the shared state of a group, hence
has access to the group's secrets.
A signed object describing a client's identity and capabilities, and including
a hybrid public-key encryption (HPKE ) public key that
can be used to encrypt to that client.
A key package that is prepublished by a client, which other clients can use to
introduce the client to a new group.
A signing key pair used to authenticate the sender of a message.Terminology specific to tree computations is described in
.We use the TLS presentation language to
describe the structure of protocol messages.This protocol is designed to execute in the context of a Service Provider (SP)
as described in . In particular, we assume
the SP provides the following services:A signature key provider which allows clients to authenticate
protocol messages in a group.A broadcast channel, for each group, which will relay a message to all members
of a group. For the most part, we assume that this channel delivers messages
in the same order to all participants. (See for further
considerations.)A directory to which clients can publish key packages and download
key packages for other participants.The goal of this protocol is to allow a group of clients to exchange
confidential and authenticated messages. It does so by deriving a sequence
of secrets and keys known only to members. Those should be secret against an
active network adversary and should have both forward secrecy and
post-compromise security with respect to compromise of any members.We describe the information stored by each client as state, which includes
both public and private data. An initial state is set up by a group creator,
which is a group containing only itself. The creator then sends Add
proposals for each client in the initial set of members, followed by a Commit
message which incorporates all of the Adds into the group state. Finally, the
group creator generates a Welcome message corresponding to the Commit and
sends this directly to all the new members, who can use the information
it contains to set up their own group state and derive a shared
secret. Members exchange Commit messages for post-compromise security, to add new
members, and to remove existing members. These messages produce new shared
secrets which are causally linked to their predecessors, forming a logical
Directed Acyclic Graph (DAG) of states.The protocol algorithms we specify here follow. Each algorithm specifies
both (i) how a client performs the operation and (ii) how other clients
update their state based on it.There are three major operations in the lifecycle of a group:Adding a member, initiated by a current member;Updating the leaf secret of a member;Removing a member.Each of these operations is "proposed" by sending a message of the corresponding
type (Add / Update / Remove). The state of the group is not changed, however,
until a Commit message is sent to provide the group with fresh entropy. In this
section, we show each proposal being committed immediately, but in more advanced
deployment cases an application might gather several proposals before
committing them all at once.Before the initialization of a group, clients publish InitKeys (as KeyPackage
objects) to a directory provided by the Service Provider.| |
| | | | |
| | KeyPackageB | | |
| |-------------------------------->| |
| | | | |
| | | KeyPackageC | |
| | |--------------->| |
| | | | |
]]>When a client A wants to establish a group with B and C, it first initializes a
group state containing only itself and downloads KeyPackages for B and C. For
each member, A generates an Add and Commit message adding that member, and
broadcasts them to the group. It also generates a Welcome message and sends this
directly to the new member (there's no need to send it to the group). Only after
A has received its Commit message back from the server does it update its state
to reflect the new member's addition.Upon receiving the Welcome message and the corresponding Commit, the new member
will be able to read and send new messages to the group. Messages received
before the client has joined the group are ignored.AB) |
| | | | Commit(Add) |
|--------------------------------------------------------------->|
| | | | |
| Welcome(B) | | | |
|------------->|state.init() | | |
| | | | |
| | | | Add(A->AB) |
| | | | Commit(Add) |
|<---------------------------------------------------------------|
|state.add(B) |<------------------------------------------------|
| |state.join() | | |
| | | | |
| | | | Add(AB->ABC) |
| | | | Commit(Add) |
|--------------------------------------------------------------->|
| | | | |
| | Welcome(C) | | |
|---------------------------->|state.init() | |
| | | | |
| | | | Add(AB->ABC) |
| | | | Commit(Add) |
|<---------------------------------------------------------------|
|state.add(C) |<------------------------------------------------|
| |state.add(C) |<---------------------------------|
| | |state.join() | |
]]>Subsequent additions of group members proceed in the same way. Any
member of the group can download a KeyPackage for a new client
and broadcast an Add message that the current group can use to update
their state, and a Welcome message that the new client can use to
initialize its state.To enforce the forward secrecy and post-compromise security of messages,
each member periodically updates their leaf secret.
Any member can update this information at any time by generating a fresh
KeyPackage and sending an Update message followed by a Commit message.
Once all members have processed both, the group's secrets will be unknown to an
attacker that had compromised the sender's prior leaf secret.Update messages should be sent at regular intervals of time as long as the group
is active, and members that don't update should eventually be removed from the
group. It's left to the application to determine an appropriate amount of time
between Updates.|
| Commit(Upd) | | | |
|---------------------------------------------------------->|
| | | | |
| | | | Update(B) |
| | | | Commit(Upd) |
|<----------------------------------------------------------|
|state.upd(B) |<-------------------------------------------|
| |state.upd(B) |<----------------------------|
| | |state.upd(B) | |
| | | | |
]]>Members are removed from the group in a similar way.
Any member of the group can send a Remove proposal followed by a
Commit message, which adds new entropy to the group state
that's known to all except the removed member.
Note that this does not necessarily imply that any member
is actually allowed to evict other members; groups can
enforce access control policies on top of these
basic mechanism.|
| | | | |
| | | | Remove(B) |
| | | | Commit(Rem) |
|<----------------------------------------------------------|
|state.rem(B) | |<----------------------------|
| | |state.rem(B) | |
| | | | |
| | | | |
]]>The protocol uses "ratchet trees" for deriving shared secrets among
a group of clients.Trees consist of nodes. A node is a
leaf if it has no children, and a parent otherwise; note that all
parents in our trees have precisely
two children, a left child and a right child. A node is the root
of a tree if it has no parents, and intermediate if it has both
children and parents. The descendants of a node are that node, its
children, and the descendants of its children, and we say a tree
contains a node if that node is a descendant of the root of the
tree. Nodes are siblings if they share the same parent.A subtree of a tree is the tree given by the descendants of any
node, the head of the subtree. The size of a tree or subtree is the
number of leaf nodes it contains. For a given parent node, its left
subtree is the subtree with its left child as head (respectively
right subtree).All trees used in this protocol are left-balanced binary trees. A
binary tree is full (and balanced) if its size is a power of
two and for any parent node in the tree, its left and right subtrees
have the same size.A binary tree is left-balanced if for every
parent, either the parent is balanced, or the left subtree of that
parent is the largest full subtree that could be constructed from
the leaves present in the parent's own subtree.
Given a list of n items, there is a unique left-balanced
binary tree structure with these elements as leaves.(Note that left-balanced binary trees are the same structure that is
used for the Merkle trees in the Certificate Transparency protocol
.)The direct path of a root is the empty list, and of any other node
is the concatenation of that node's parent along with the parent's direct path.
The copath of a node is the node's sibling concatenated with the list of
siblings of all the nodes in its direct path, excluding the root.For example, in the below tree:The direct path of C is (CD, ABCD, ABCDEFG)The copath of C is (D, AB, EFG)Each node in the tree is assigned a node index, starting at zero and
running from left to right. A node is a leaf node if and only if it
has an even index. The node indices for the nodes in the above tree
are as follows:0 = A1 = AB2 = B3 = ABCD4 = C5 = CD6 = D7 = ABCDEFG8 = E9 = EF10 = F11 = EFG12 = GThe leaves of the tree are indexed separately, using a leaf index,
since the protocol messages only need to refer to leaves in the
tree. Like nodes, leaves are numbered left to right. The node with
leaf index k is also called the k-th leaf. Note that
given the above numbering, a node is a leaf node if and only if it
has an even node index, and a leaf node's leaf index is half its
node index. The leaf indices in the above tree are as follows:0 = A1 = B2 = C3 = D4 = E5 = F6 = GA particular instance of a ratchet tree is defined by the same parameters that
define an instance of HPKE, namely:A Key Encapsulation Mechanism (KEM), including a DeriveKeyPair function that
creates a key pair for the KEM from a symmetric secretA Key Derivation Function (KDF), including Extract and Expand functionsAn AEAD encryption schemeEach node in a ratchet tree contains up to five values:A private key (only within the member's direct path, see below)A public keyAn ordered list of leaf indices for "unmerged" leaves (see
)A credential (only for leaf nodes)A hash of the node's parent, as of the last time the node was changed.The conditions under which each of these values must or must not be
present are laid out in .A node in the tree may also be blank, indicating that no value is
present at that node. The resolution of a node is an ordered list
of non-blank nodes that collectively cover all non-blank descendants
of the node.The resolution of a non-blank node comprises the node itself,
followed by its list of unmerged leaves, if anyThe resolution of a blank leaf node is the empty listThe resolution of a blank intermediate node is the result of
concatenating the resolution of its left child with the resolution
of its right child, in that orderFor example, consider the following tree, where the "_" character
represents a blank node:In this tree, we can see all of the above rules in play:The resolution of node 5 is the list [CD, C]The resolution of node 2 is the empty list []The resolution of node 3 is the list [A, CD, C]Every node, regardless of whether the node is blank or populated, has
a corresponding hash that summarizes the contents of the subtree
below that node. The rules for computing these hashes are described
in .We generally assume that each participant maintains a complete and
up-to-date view of the public state of the group's ratchet tree,
including the public keys for all nodes and the credentials
associated with the leaf nodes.No participant in an MLS group knows the private key associated with
every node in the tree. Instead, each member is assigned to a leaf of the tree,
which determines the subset of private keys it knows. The
credential stored at that leaf is one provided by the member.In particular, MLS maintains the members' views of the tree in such
a way as to maintain the tree invariant:In other words, if a node is not blank, then it holds a public key.
The corresponding private key is known only to members occupying
leaves below that node.The reverse implication is not true: A member may not know the private keys of
all the intermediate nodes they're below. Such a member has an unmerged leaf.
Encrypting to an intermediate node requires encrypting to the node's public key,
as well as the public keys of all the unmerged leaves below it. A leaf is
unmerged when it is first added, because the process of adding the leaf does not
give it access to all of the nodes above it in the tree. Leaves are "merged" as
they receive the private keys for nodes, as described in
.A member of an MLS group advances the key schedule to provide forward secrecy
and post-compromise security by providing the group with fresh key material to
be added into the group's shared secret.
To do so, one member of the group generates fresh key
material, applies it to their local tree state, and then sends this key material
to other members in the group via an UpdatePath message (see ) .
All other group members then apply the key material in the UpdatePath to their
own local tree state to derive the group's now-updated shared secret.To begin, the generator of the UpdatePath updates its leaf
KeyPackage and its direct path to the root with new secret values. The
HPKE leaf public key within the KeyPackage MUST be derived from a freshly
generated HPKE secret key to provide post-compromise security.The generator of the UpdatePath starts by sampling a fresh random value called
"leaf_secret", and uses the leaf_secret to generate their leaf HPKE key pair
(see ) and to seed a sequence of "path secrets", one for each
ancestor of its leaf. In this setting,
path_secret[0] refers to the node directly above the leaf,
path_secret[1] for its parent, and so on. At each step, the path
secret is used to derive a new secret value for the corresponding
node, from which the node's key pair is derived.For example, suppose there is a group with four members:If member B subsequently generates an UpdatePath based on a secret
"leaf_secret", then it would generate the following sequence
of path secrets: node_secret[1] --> node_priv[1], node_pub[1]
^
|
path_secret[0] --> node_secret[0] --> node_priv[0], node_pub[0]
^
|
leaf_secret --> leaf_node_secret --> leaf_priv, leaf_pub
~> leaf_key_package
]]>After applying the UpdatePath, the tree will have the following structure, where
"np[i]" represents the node_priv values generated as described
above:After performing these operations, the generator of the UpdatePath MUST
delete the leaf_secret.After generating fresh key material and applying it to ratchet forward their
local tree state as described in the prior section, the generator must broadcast
this update to other members of the group in a Commit message, who
apply it to keep their local views of the tree in
sync with the sender's. More specifically, when a member commits a change to
the tree (e.g., to add or remove a member), it transmits an UpdatePath message
containing a set of public and encrypted private
values for intermediate nodes in the direct path of a leaf. The
other members of the group use these values to update
their view of the tree, aligning their copy of the tree to the
sender's.To transmit this update, the sender broadcasts to the group
the following information for each node in the direct path of the
leaf, including the root:The public key for the nodeZero or more encrypted copies of the path secret corresponding to
the nodeThe path secret value for a given node is encrypted for the subtree
corresponding to the parent's non-updated child, that is, the child
on the copath of the leaf node.
There is one encrypted path secret for each public key in the resolution
of the non-updated child.The recipient of a path update processes it with the following steps:Compute the updated path secrets.
Identify a node in the direct path for which the local member
is in the subtree of the non-updated child.Identify a node in the resolution of the copath node for
which this node has a private key.Decrypt the path secret for the parent of the copath node using
the private key from the resolution node.Derive path secrets for ancestors of that node using the
algorithm described above.The recipient SHOULD verify that the received public keys agree
with the public keys derived from the new path_secret values.Merge the updated path secrets into the tree.
For all updated nodes,
Replace the public key for each node with the received public key.Set the list of unmerged leaves to the empty list.Store the updated hash of the node's parent (represented as a ParentNode
struct), going from root to leaf, so that each hash incorporates all the
nodes above it. The root node always has a zero-length hash for this
value.For nodes where an updated path secret was computed in step 1,
compute the corresponding node key pair and replace the values
stored at the node with the computed values.For example, in order to communicate the example update described in
the previous section, the sender would transmit the following
values:Public KeyCiphertext(s)pk(ns[1])E(pk(C), ps[1]), E(pk(D), ps[1])pk(ns[0])E(pk(A), ps[0])In this table, the value pk(ns[X]) represents the public key
derived from the node secret X, whereas pk(X) represents the public leaf key
for user X. The value E(K, S) represents
the public-key encryption of the path secret S to the
public key K.After processing the update, each recipient MUST delete outdated key material,
specifically:The path secrets used to derive each updated node key pair.Each outdated node key pair that was replaced by the update.Each MLS session uses a single ciphersuite that specifies the
following primitives to be used in group key computations:HPKE parameters:
A Key Encapsulation Mechanism (KEM)A Key Derivation Function (KDF)An AEAD encryption algorithmA hash algorithmA signature algorithmThe HPKE parameters are used to instantiate HPKE for the
purpose of public-key encryption. The DeriveKeyPair function associated to
the KEM for the ciphersuite maps octet strings to HPKE key pairs.Ciphersuites are represented with the CipherSuite type. HPKE public keys
are opaque values in a format defined by the underlying
protocol (see the Cryptographic Dependencies section of the HPKE specification for more
information).;
]]>The signature algorithm specified in the ciphersuite is the mandatory algorithm
to be used for signatures in MLSPlaintext and the tree signatures. It MUST be
the same as the signature algorithm specified in the credential field of the
KeyPackage objects in the leaves of the tree (including the InitKeys
used to add new members).The ciphersuites are defined in section .A member of a group authenticates the identities of other participants by means
of credentials issued by some authentication system, like a PKI. Each type of
credential MUST express the following data in the context of the group it is
used with:The public key of a signature key pair matching the SignatureScheme specified
by the CipherSuite of the groupThe identity of the holder of the private keyCredentials MAY also include information that allows a relying party
to verify the identity / signing key binding.Additionally, Credentials SHOULD specify the signature scheme corresponding to
each contained public key.;
SignatureScheme signature_scheme;
opaque signature_key<0..2^16-1>;
} BasicCredential;
struct {
opaque cert_data<0..2^16-1>;
} Certificate;
struct {
CredentialType credential_type;
select (Credential.credential_type) {
case basic:
BasicCredential;
case x509:
Certificate chain<1..2^32-1>;
};
} Credential;
]]>A BasicCredential is a raw, unauthenticated assertion of an identity/key
binding. The format of the key in the public_key field is defined by the
relevant ciphersuite: the group ciphersuite for a credential in a ratchet tree,
the KeyPackage ciphersuite for a credential in a KeyPackage object.For X509Credential, each entry in the chain represents a single DER-encoded
X509 certificate. The chain is ordered such that the first entry (chain[0])
is the end-entity certificate and each subsequent certificate in the chain
MUST be the issuer of the previous certificate. The algorithm for the
public_key in the end-entity certificate MUST match the relevant
ciphersuite.For ciphersuites using Ed25519 or Ed448 signature schemes, the public key is in
the format specified . For ciphersuites using ECDSA with the NIST
curves P-256 or P-521, the public key is the output of the uncompressed
Elliptic-Curve-Point-to-Octet-String conversion according to .Note that each new credential that has not already been validated
by the application MUST be validated against the Authentication
Service.In order to facilitate asynchronous addition of clients to a
group, it is possible to pre-publish key packages that
provide some public information about a user. KeyPackage
structures provide information about a client that any existing
member can use to add this client to the group asynchronously.A KeyPackage object specifies a ciphersuite that the client
supports, as well as providing a public key that others can use
for key agreement. The client's signature key can be updated
throughout the lifetime of the group by sending a new KeyPackage
with a new Credential. However, the identity MUST be the same in
both Credentials and the new Credential MUST be validated by the
authentication service.When used as InitKeys, KeyPackages are intended to be used only once and SHOULD NOT
be reused except in case of last resort. (See ).
Clients MAY generate and publish multiple InitKeys to
support multiple ciphersuites.KeyPackages contain a public key chosen by the client, which the
client MUST ensure uniquely identifies a given KeyPackage object
among the set of KeyPackages created by this client.The value for hpke_init_key MUST be a public key for the asymmetric
encryption scheme defined by cipher_suite. The whole structure
is signed using the client's signature key. A KeyPackage object
with an invalid signature field MUST be considered malformed.
The input to the signature computation comprises all of the fields
except for the signature field.;
} Extension;
struct {
ProtocolVersion version;
CipherSuite cipher_suite;
HPKEPublicKey hpke_init_key;
Credential credential;
Extension extensions<8..2^32-1>;
opaque signature<0..2^16-1>;
} KeyPackage;
]]>KeyPackage objects MUST contain at least two extensions, one of type
capabilities, and one of
type lifetime. The capabilities extension
allow MLS session establishment to be safe from downgrade attacks on the
parameters described (as discussed in ), while still only advertising
one version / ciphersuite per KeyPackage.As the KeyPackage is a structure which is stored in the Ratchet
Tree and updated depending on the evolution of this tree, each
modification of its content MUST be reflected by a change of its
signature. This allow other members to control the validity of the KeyPackage
at any time and in particular in the case of a newcomer joining the group.The capabilities extension indicates what protocol versions, ciphersuites, and
protocol extensions are supported by a client.;
CipherSuite ciphersuites<0..255>;
ExtensionType extensions<0..255>;
} Capabilities;
]]>This extension MUST be always present in a KeyPackage. Extensions that appear
in the extensions field of a KeyPackage MUST be included in the extensions
field of the capabilities extension.The lifetime extension represents the times between which clients will
consider a KeyPackage valid. This time is represented as an absolute time,
measured in seconds since the Unix epoch (1970-01-01T00:00:00Z). A client MUST
NOT use the data in a KeyPackage for any processing before the not_before
date, or after the not_after date.Applications MUST define a maximum total lifetime that is acceptable for a
KeyPackage, and reject any KeyPackage where the total lifetime is longer than
this duration.This extension MUST always be present in a KeyPackage.Within MLS, a KeyPackage is identified by its hash (see, e.g.,
). The key_id extension allows applications to add
an explicit, application-defined identifier to a KeyPackage.;
]]>The parent_hash extension serves to bind a KeyPackage to all the nodes
above it in the group's ratchet tree. This enforces the tree invariant, meaning
that malicious members can't lie about the state of the ratchet tree when they
send Welcome messages to new members.;
]]>This extension MUST be present in all Updates that are sent as part of a Commit
message. If the extension is present, clients MUST verify that parent_hash
matches the hash of the leaf's parent node when represented as a ParentNode
struct.To allow group members to verify that they agree on the public
cryptographic state of the group, this section defines a scheme for
generating a hash value that represents the contents of the group's
ratchet tree and the members' KeyPackages.The hash of a tree is the hash of its root node, which we define
recursively, starting with the leaves.Elements of the ratchet tree are called Node objects and
the leaves contain an optional KeyPackage, while the parents contain
an optional ParentNode.;
struct {
HPKEPublicKey public_key;
uint32 unmerged_leaves<0..2^32-1>;
opaque parent_hash<0..255>;
} ParentNode;
]]>When computing the hash of a parent node, the ParentNodeHashInput
structure is used: parent_node;
opaque left_hash<0..255>;
opaque right_hash<0..255>;
} ParentNodeHashInput;
]]>The left_hash and right_hash fields hold the hashes of the node's
left and right children, respectively. When computing the hash of
a leaf node, the hash of a LeafNodeHashInput object is used: key_package;
} LeafNodeHashInput;
]]>Note that the node_index field contains the index of the leaf among the nodes
in the tree, not its index among the leaves; node_index = 2 * leaf_index.Each member of the group maintains a GroupContext object that
summarizes the state of the group:;
uint64 epoch;
opaque tree_hash<0..255>;
opaque confirmed_transcript_hash<0..255>;
Extension extensions<0..2^32-1>;
} GroupContext;
]]>The fields in this state have the following semantics:The group_id field is an application-defined identifier for the
group.The epoch field represents the current version of the group key.The tree_hash field contains a commitment to the contents of the
group's ratchet tree and the credentials for the members of the
group, as described in .The confirmed_transcript_hash field contains a running hash over
the messages that led to this state.When a new member is added to the group, an existing member of the
group provides the new member with a Welcome message. The Welcome
message provides the information the new member needs to initialize
its GroupContext.Different changes to the group will have different effects on the group state.
These effects are described in their respective subsections of .
The following general rules apply:The group_id field is constantThe epoch field increments by one for each Commit message that
is processedThe tree_hash is updated to represent the current tree and
credentialsThe confirmed_transcript_hash is updated with the data for an
MLSPlaintext message encoding a Commit message in two parts:;
uint64 epoch;
Sender sender;
ContentType content_type = commit;
Commit commit;
opaque signature<0..2^16-1>;
} MLSPlaintextCommitContent;
struct {
MAC confirmation_tag;
} MLSPlaintextCommitAuthData;
interim_transcript_hash_[0] = ""; // zero-length octet string
confirmed_transcript_hash_[n] =
Hash(interim_transcript_hash_[n] ||
MLSPlaintextCommitContent_[n]);
interim_transcript_hash_[n+1] =
Hash(confirmed_transcript_hash_[n] ||
MLSPlaintextCommitAuthData_[n]);
]]>Thus the confirmed_transcript_hash field in a GroupContext object represents a
transcript over the whole history of MLSPlaintext Commit messages, up to the
confirmation tag field in the current MLSPlaintext message. The confirmation tag
is then included in the transcript for the next epoch. The interim transcript
hash is passed to new members in the GroupInfo struct, and enables existing
members to incorporate a Commit message into the transcript without having to
store the whole MLSPlaintextCommitAuthData structure.As shown above, when a new group is created, the interim_transcript_hash field
is set to the zero-length octet string.As described in , each MLS Commit message may optionally
transmit a KeyPackage leaf and node values along its direct path.
The path contains a public key and encrypted secret value for all
intermediate nodes in the path above the leaf. The path is ordered
from the closest node to the leaf to the root; each node MUST be the
parent of its predecessor.;
opaque ciphertext<0..2^16-1>;
} HPKECiphertext;
struct {
HPKEPublicKey public_key;
HPKECiphertext encrypted_path_secret<0..2^32-1>;
} UpdatePathNode;
struct {
KeyPackage leaf_key_package;
UpdatePathNode nodes<0..2^32-1>;
} UpdatePath;
]]>For each UpdatePathNode, the resolution of the corresponding copath node MUST
be filtered by removing all new leaf nodes added as part of this MLS Commit
message. The number of ciphertexts in the encrypted_path_secret vector MUST be
equal to the length of the filtered resolution, with each ciphertext being the
encryption to the respective resolution node.The HPKECiphertext values are computed aswhere node_public_key is the public key of the node that the path
secret is being encrypted for, group_context is the current GroupContext object
for the group, and the functions SetupBaseS and
Seal are defined according to .Decryption is performed in the corresponding way, using the private
key of the resolution node and the ephemeral public key
transmitted in the message.Group keys are derived using the Extract and Expand functions from the KDF
for the group's ciphersuite, as well as the functions defined below: = "mls10 " + Label;
opaque context<0..2^32-1> = Context;
} KDFLabel;
DeriveSecret(Secret, Label) =
ExpandWithLabel(Secret, Label, "", KDF.Nh)
]]>The value KDF.Nh is the size of an output from KDF.Extract, in bytes. In
the below diagram:KDF.Extract takes its salt argument from the top and its IKM
argument from the leftDeriveSecret takes its Secret argument from the incoming arrowWhen processing a handshake message, a client combines the
following information to derive new epoch secrets:The init secret from the previous epochThe commit secret for the current epochThe GroupContext object for current epochGiven these inputs, the derivation of secrets for an epoch
proceeds as shown in the following diagram: KDF.Extract = joiner_secret
|
V
Derive-Secret(., "member")
|
V
psk_secret (or 0) -> KDF.Extract = member_secret
|
+--> Derive-Secret(., "welcome")
| = welcome_secret
|
V
ExpandWithLabel(., "epoch", GroupContext_[n], KDF.Nh)
|
V
epoch_secret
|
+--> Derive-Secret(., A number of secrets are derived from the epoch secret for different purposes:SecretLabelsender_data_secret"sender data"encryption_secret"encryption"exporter_secret"exporter"authentication_secret"authentication"external_secret"external"confirmation_key"confirm"membership_key"membership"resumption_secret"resumption"The "external secret" is used to derive an HPKE key pair whose private key is
held by the entire group:The public key external_pub can be published as part of the PublicGroupState
struct in order to allow non-members to join the group using an external commit.In addition to initializing a new epoch via KDF invocations as described above,
an MLS group can also initialize a new epoch via an asymmetric interaction using
the external key pair for the previous epoch. This is done when an new member
is joining via an external commit.In this process, the joiner sends a new init_secret value to the group using
the HPKE export method. The joiner then uses that init_secret with
information provided in the PublicGroupState and an external Commit to initialize
their copy of the key schedule for the new epoch.Members of the group receive the kem_output in an ExternalInit proposal and
preform the corresponding calculation to retrieve the init_secret value.In both cases, the info input to HPKE is set to the PublicGroupState for the
previous epoch, encoded using the TLS serialization.Groups which already have an out-of-band mechanism to generate
shared group secrets can inject those into the MLS key schedule to seed
the MLS group secrets computations by this external entropy.Injecting an external PSK can improve security in the case
where having a full run of updates across members is too expensive, or if
the external group key establishment mechanism provides
stronger security against classical or quantum adversaries.Note that, as a PSK may have a different lifetime than an update, it
does not necessarily provide the same Forward Secrecy (FS) or Post-Compromise
Security (PCS) guarantees as a Commit message.Each PSK in MLS has a type that designates how it was provisioned.
External PSKs are provided by the application, while recovery and re-init PSKs
are derived from the MLS key schedule and used in cases where it is
necessary to authenticate a member's participation in a prior group state.
In particular, in addition to external PSK types, a PSK derived from within MLS
may be used in the following cases:Re-Initialization: If during the lifetime of a group, the group members
decide to switch to a more secure ciphersuite or newer protocol version,
a PSK can be used to carry entropy from the old group forward into a new
group with the desired parameters.Branching: A PSK may be used to bootstrap a subset of current group
members into a new group. This applies if a subset of current group
members wish to branch based on the current group state.The injection of one or more PSKs into the key schedule is signaled in two ways:
1) as a PreSharedKey proposal, and 2) in the GroupSecrets object of a
Welcome message sent to new members added in that epoch.;
case reinit:
opaque psk_group_id<0..255>;
uint64 psk_epoch;
case branch:
opaque psk_group_id<0..255>;
uint64 psk_epoch;
}
opaque psk_nonce<0..255>;
} PreSharedKeyID;
struct {
PreSharedKeyID psks<0..2^16-1>;
} PreSharedKeys;
]]>On receiving a Commit with a PreSharedKey proposal or a GroupSecrets object
with the psks field set, the receiving Client includes them in the key
schedule in the order listed in the Commit, or in the psks field respectively.
For resumption PSKs, the PSK is defined as the resumption_secret of the group and
epoch specified in the PreSharedKeyID object. Specifically, psk_secret is
computed as follows:The index field in PSKLabel corresponds to the index of the PSK in the psk
array, while the count field contains the total number of PSKs.For the generation of encryption keys and nonces, the key schedule begins with
the encryption_secret at the root and derives a tree of secrets with the same
structure as the group's ratchet tree. Each leaf in the Secret Tree is
associated with the same group member as the corresponding leaf in the ratchet
tree. Nodes are also assigned an index according to their position in the array
representation of the tree (described in ). If N is a node index in
the Secret Tree then left(N) and right(N) denote the children of N (if they
exist).The secret of any other node in the tree is derived from its parent's secret
using a call to DeriveTreeSecret:If N is a node index in the Secret Tree then the secrets of the children
of N are defined to be: DeriveTreeSecret(., "tree", left(N), 0, KDF.Nh)
| = tree_node_[left(N)]_secret
|
+--> DeriveTreeSecret(., "tree", right(N), 0, KDF.Nh)
= tree_node_[right(N)]_secret
]]>The secret in the leaf of the Secret Tree is used to initiate two symmetric hash
ratchets, from which a sequence of single-use keys and nonces are derived, as
described in . The root of each ratchet is computed as: DeriveTreeSecret(., "handshake", N, 0, KDF.Nh)
| = handshake_ratchet_secret_[N]_[0]
|
+--> DeriveTreeSecret(., "application", N, 0, KDF.Nh)
= application_ratchet_secret_[N]_[0]
]]>As described in , MLS encrypts three different
types of information:Metadata (sender information)Handshake messages (Proposal and Commit)Application messagesThe sender information used to look up the key for content encryption is
encrypted with an AEAD where the key and nonce are derived from both
sender_data_secret and a sample of the encrypted message content.For handshake and application messages, a sequence of keys is derived via a
"sender ratchet". Each sender has their own sender ratchet, and each step along
the ratchet is called a "generation".A sender ratchet starts from a per-sender base secret derived from a Secret
Tree, as described in . The base secret initiates a symmetric
hash ratchet which generates a sequence of keys and nonces. The sender uses the
j-th key/nonce pair in the sequence to encrypt (using the AEAD) the j-th message
they send during that epoch. Each key/nonce pair MUST NOT be used to encrypt
more than one message.Keys, nonces, and the secrets in ratchets are derived using
DeriveTreeSecret. The context in a given call consists of the index
of the sender's leaf in the ratchet tree and the current position in
the ratchet. In particular, the node index of the sender's leaf in the
ratchet tree is the same as the node index of the leaf in the Secret Tree
used to initialize the sender's ratchet. DeriveTreeSecret(., "nonce", N, j, AEAD.Nn)
| = ratchet_nonce_[N]_[j]
|
+--> DeriveTreeSecret(., "key", N, j, AEAD.Nk)
| = ratchet_key_[N]_[j]
|
V
DeriveTreeSecret(., "secret", N, j, KDF.Nh)
= ratchet_secret_[N]_[j+1]
]]>Here, AEAD.Nn and AEAD.Nk denote the lengths
in bytes of the nonce and key for the AEAD scheme defined by
the ciphersuite.It is important to delete all security-sensitive values as soon as they are
consumed. A sensitive value S is said to be consumed ifS was used to encrypt or (successfully) decrypt a message, or ifa key, nonce, or secret derived from S has been consumed. (This goes for
values derived via DeriveSecret as well as ExpandWithLabel.)Here, S may be the init_secret, commit_secret, epoch_secret,
encryption_secret as well as any secret in a Secret Tree or one of the
ratchets.As soon as a group member consumes a value they MUST immediately delete
(all representations of) that value. This is crucial to ensuring
forward secrecy for past messages. Members MAY keep unconsumed values around
for some reasonable amount of time to handle out-of-order message delivery.For example, suppose a group member encrypts or (successfully) decrypts an
application message using the j-th key and nonce in the ratchet of node
index N in some epoch n. Then, for that member, at least the following
values have been consumed and MUST be deleted:the commit_secret, joiner_secret, member_secret, epoch_secret,
encryption_secret of that epoch n as well as the init_secret of the
previous epoch n-1,all node secrets in the Secret Tree on the path from the root to the leaf with
node index N,the first j secrets in the application data ratchet of node index N andapplication_ratchet_nonce_[N]_[j] and application_ratchet_key_[N]_[j].Concretely, suppose we have the following Secret Tree and ratchet for
participant D:Then if a client uses key K1 and nonce N1 during epoch n then it must consume
(at least) values G, F, D, AR0, AR1, K1, N1 as well as the key schedule secrets
used to derive G (the encryption_secret), namely init_secret of epoch n-1
and commit_secret, joiner_secret, member_secret, epoch_secret of epoch
n. The client MAY retain (not consume) the values K0 and N0 to allow for
out-of-order delivery, and SHOULD retain AR2 for processing future messages.The main MLS key schedule provides an exporter_secret which can
be used by an application as the basis to derive new secrets called
exported_value outside the MLS layer.Each application SHOULD provide a unique label to MLS-Exporter that
identifies its use case. This is to prevent two
exported outputs from being generated with the same values and used
for different functionalities.The exported values are bound to the group epoch from which the
exporter_secret is derived, hence reflects a particular state of
the group.It is RECOMMENDED for the application generating exported values
to refresh those values after a Commit is processed.The main MLS key schedule provides a resumption_secret which can provide extra
security in some cross-group operations.The application SHOULD specify an upper limit on the number of past
epochs for which the resumption_secret may be stored.There are two ways in which a resumption_secret can be used: to re-initialize
the group with different parameters, or to create a
sub-group of an existing group as detailed in .Resumption keys are distinguished from exporter keys in that they have specific
use inside the MLS protocol, whereas the use of exporter secrets may be
decided by an external application. They are thus derived separately to avoid
key material reuse.The main MLS key schedule provides a per-epoch authentication_secret.
If one of the parties is being actively impersonated by an attacker, their
authentication_secret will differ from that of the other group members.
Thus, members of a group MAY use their authentication_secrets within
an out-of-band authentication protocol to ensure that they
share the same view of the group.Handshake and application messages use a common framing structure.
This framing provides encryption to ensure confidentiality within the
group, as well as signing to authenticate the sender within the group.The two main structures involved are MLSPlaintext and MLSCiphertext.
MLSCiphertext represents a signed and encrypted message, with
protections for both the content of the message and related
metadata. MLSPlaintext represents a message that is only signed,
and not encrypted. Applications MUST use MLSCiphertext to encrypt
application messages and SHOULD use MLSCiphertext to encode
handshake messages, but MAY transmit handshake messages encoded
as MLSPlaintext objects in cases where it is necessary for the
Delivery Service to examine such messages.;
} MAC;
struct {
opaque group_id<0..255>;
uint64 epoch;
Sender sender;
opaque authenticated_data<0..2^32-1>;
ContentType content_type;
select (MLSPlaintext.content_type) {
case application:
opaque application_data<0..2^32-1>;
case proposal:
Proposal proposal;
case commit:
Commit commit;
}
opaque signature<0..2^16-1>;
optional confirmation_tag;
optional membership_tag;
} MLSPlaintext;
struct {
opaque group_id<0..255>;
uint64 epoch;
ContentType content_type;
opaque authenticated_data<0..2^32-1>;
opaque encrypted_sender_data<0..255>;
opaque ciphertext<0..2^32-1>;
} MLSCiphertext;
]]>The field confirmation_tag MUST be present if content_type equals commit.
Otherwise, it MUST NOT be present.External sender types are sent as MLSPlaintext, see
for their use.The remainder of this section describes how to compute the signature of an
MLSPlaintext object and how to convert it to an MLSCiphertext object for
member sender types. The steps are:Set group_id, epoch, content_type and authenticated_data fields from the
MLSPlaintext object directlyIdentify the key and key generation depending on the content typeEncrypt an MLSCiphertextContent for the ciphertext field using the key
identified and MLSPlaintext objectEncrypt the sender data using a key and nonce derived from the
sender_data_secret for the epoch and a sample of the encrypted
MLSCiphertextContent.Decryption is done by decrypting the sender data, then the message, and then
verifying the content signature.The following sections describe the encryption and signing processes in detail.The signature field in an MLSPlaintext object is computed using the signing
private key corresponding to the credential at the leaf of the tree indicated by
the sender field. The signature covers the plaintext metadata and message
content, which is all of MLSPlaintext except for the signature, the
confirmation_tag and membership_tag fields. If the sender is a member of the
group, the signature also covers the GroupContext for the current epoch, so that
signatures are specific to a given group and epoch.;
uint64 epoch;
Sender sender;
opaque authenticated_data<0..2^32-1>;
ContentType content_type;
select (MLSPlaintextTBS.content_type) {
case application:
opaque application_data<0..2^32-1>;
case proposal:
Proposal proposal;
case commit:
Commit commit;
}
} MLSPlaintextTBS;
]]>The membership_tag field in the MLSPlaintext object authenticates the sender's
membership in the group. For an MLSPlaintext with a sender type other than
member, this field MUST be omitted. For messages sent by members, it MUST be
present and set to the following value:;
optional confirmation_tag;
} MLSPlaintextTBM;
membership_tag = MAC(membership_key, MLSPlaintextTBM);
]]>Note that the membership_tag only needs to be computed for MLSPlaintext
messages that will be sent over the wire, and isn't needed for those that will
be encrypted and transmitted as MLSCiphertext messages.The ciphertext field of the MLSCiphertext object is produced by supplying the
inputs described below to the AEAD function specified by the ciphersuite in use.
The plaintext input contains the content and signature of the MLSPlaintext, plus
optional padding. These values are encoded in the following form:;
case proposal:
Proposal proposal;
case commit:
Commit commit;
}
opaque signature<0..2^16-1>;
optional confirmation_tag;
opaque padding<0..2^16-1>;
} MLSCiphertextContent;
]]>In the MLS key schedule, the sender creates two distinct key ratchets for
handshake and application messages for each member of the group. When encrypting
a message, the sender looks at the ratchets it derived for its own member and
chooses an unused generation from either the handshake or application ratchet
depending on the content type of the message. This generation of the ratchet is
used to derive a provisional nonce and key.Before use in the encryption operation, the nonce is XORed with a fresh random
value to guard against reuse. Because the key schedule generates nonces
deterministically, a client must keep persistent state as to where in the key
schedule it is; if this persistent state is lost or corrupted, a client might
reuse a generation that has already been used, causing reuse of a key/nonce pair.To avoid this situation, the sender of a message MUST generate a fresh random
4-byte "reuse guard" value and XOR it with the first four bytes of the nonce
from the key schedule before using the nonce for encryption. The sender MUST
include the reuse guard in the reuse_guard field of the sender data object, so
that the recipient of the message can use it to compute the nonce to be used for
decryption.The Additional Authenticated Data (AAD) input to the encryption
contains an object of the following form, with the values used to
identify the key and nonce:;
uint64 epoch;
ContentType content_type;
opaque authenticated_data<0..2^32-1>;
} MLSCiphertextContentAAD;
]]>The "sender data" used to look up the key for the content encryption is
encrypted with the ciphersuite's AEAD with a key and nonce derived from both the
sender_data_secret and a sample of the encrypted content. Before being
encrypted, the sender data is encoded as an object of the following form:MLSSenderData.sender is assumed to be a member sender type. When constructing
an MLSSenderData from a Sender object, the sender MUST verify Sender.sender_type
is member and use Sender.sender for MLSSenderData.sender.The reuse_guard field contains a fresh random value used to avoid nonce reuse
in the case of state loss or corruption, as described in .The key and nonce provided to the AEAD are computed as the KDF of the first
KDF.Nh bytes of the ciphertext generated in the previous section. If the
length of the ciphertext is less than KDF.Nh, the whole ciphertext is used
without padding. In pseudocode, the key and nonce are derived as:The Additional Authenticated Data (AAD) for the SenderData ciphertext is all the
fields of MLSCiphertext excluding encrypted_sender_data:;
uint64 epoch;
ContentType content_type;
} MLSSenderDataAAD;
]]>When parsing a SenderData struct as part of message decryption, the
recipient MUST verify that the sender field represents an occupied
leaf in the ratchet tree. In particular, the sender index value
MUST be less than the number of leaves in the tree.A group is always created with a single member, the "creator". The other
members are added when the creator effectively sends itself an Add proposal and
commits it, then sends the corresponding Welcome message to the new
participants. These processes are described in detail in , ,
and .The creator of a group MUST take the following steps to initialize the group:Fetch KeyPackages for the members to be added, and select a version and
ciphersuite according to the capabilities of the members. To protect against
downgrade attacks, the creator MUST use the capabilities extensions
in these KeyPackages to verify that the
chosen version and ciphersuite is the best option supported by all members.Initialize a one-member group with the following initial values (where "0"
represents an all-zero vector of size KDF.Nh):
Ratchet tree: A tree with a single node, a leaf containing an HPKE public
key and credential for the creatorGroup ID: A value set by the creatorEpoch: 0Tree hash: The root hash of the above ratchet treeConfirmed transcript hash: 0Interim transcript hash: 0Init secret: a fresh random value of size KDF.NhFor each member, construct an Add proposal from the KeyPackage for that
member (see )Construct a Commit message that commits all of the Add proposals, in any order
chosen by the creator (see )Process the Commit message to obtain a new group state (for the epoch in which
the new members are added) and a Welcome messageTransmit the Welcome message to the other new membersThe recipient of a Welcome message processes it as described in
.In principle, the above process could be streamlined by having the
creator directly create a tree and choose a random value for first
epoch's epoch secret. We follow the steps above because it removes
unnecessary choices, by which, for example, bad randomness could be
introduced. The only choices the creator makes here are its own
KeyPackage, the leaf secret from which the Commit is built, and the
intermediate key pairs along the direct path to the root.A new group may be tied to an already existing group for the purpose of
re-initializing the existing group, or to branch into a sub-group.
Re-initializing an existing group may be used, for example, to restart the group
with a different ciphersuite or protocol version. Branching may be used to
bootstrap a new group consisting of a subset of current group members, based on
the current group state.In both cases, the psk_nonce included in the PreSharedKeyID object must be a
randomly sampled nonce of length KDF.Nh to avoid key re-use.If a client wants to create a subgroup of an existing group, they MAY choose to
include a PreSharedKeyID in the GroupSecrets object of the Welcome message choosing
the psktypebranch, the group_id of the group from which a subgroup is to
be branched, as well as an epoch within the number of epochs for which a
resumption_secret is kept.Over the lifetime of a group, its membership can change, and existing members
might want to change their keys in order to achieve post-compromise security. In
MLS, each such change is accomplished by a two-step process:A proposal to make the change is broadcast to the group in a Proposal messageA member of the group or a new member broadcasts a Commit message that causes one or more
proposed changes to enter into effectThe group thus evolves from one cryptographic state to another each time a
Commit message is sent and processed. These states are referred to as "epochs"
and are uniquely identified among states of the group by eight-octet epoch values.
When a new group is initialized, its initial state epoch is 0x0000000000000000. Each time
a state transition occurs, the epoch number is incremented by one.Proposals are included in an MLSPlaintext by way of a Proposal structure that
indicates their type:On receiving an MLSPlaintext containing a Proposal, a client MUST verify the
signature on the enclosing MLSPlaintext. If the signature verifies
successfully, then the Proposal should be cached in such a way that it can be
retrieved by hash (as a ProposalOrRef object) in a later Commit message.An Add proposal requests that a client with a specified KeyPackage be added
to the group.The proposer of the Add does not control where in the group's ratchet tree the
new member is added. Instead, the sender of the Commit message chooses a
location for each added member and states it in the Commit message.An Add is applied after being included in a Commit message. The position of the
Add in the list of proposals determines the leaf index index where the new member
will be added. For the first Add in the Commit, index is the leftmost empty
leaf in the tree, for the second Add, the next empty leaf to the right, etc.If necessary, extend the tree to the right until it has at least index + 1
leavesFor each non-blank intermediate node along the path from the leaf at position
index to the root, add index to the unmerged_leaves list for the node.Set the leaf node in the tree at position index to a new node containing the
public key from the KeyPackage in the Add, as well as the credential under
which the KeyPackage was signedAn Update proposal is a similar mechanism to Add with the distinction
that it is the sender's leaf KeyPackage in the tree which would be
updated with a new KeyPackage.A member of the group applies an Update message by taking the following steps:Replace the sender's leaf KeyPackage with the one contained in
the Update proposalBlank the intermediate nodes along the path from the sender's leaf to the rootA Remove proposal requests that the client at a specified index in the tree be
removed from the group.A member of the group applies a Remove message by taking the following steps:Replace the leaf node at position removed with a blank nodeBlank the intermediate nodes along the path from the removed leaf to the rootA PreSharedKey proposal can be used to request that a pre-shared key be
injected into the key schedule in the process of advancing the epoch.The psktype of the pre-shared key MUST be external and the psk_nonce MUST
be a randomly sampled nonce of length KDF.Nh. When processing a Commit message
that includes one or more PreSharedKey proposals, group members derive
psk_secret as described in , where the order of the PSKs
corresponds to the order of the PreSharedKey proposals in the Commit.A ReInit proposal represents a request to re-initialize the group with different
parameters, for example, to increase the version number or to change the
ciphersuite. The re-initialization is done by creating a completely new group
and shutting down the old one.;
ProtocolVersion version;
CipherSuite cipher_suite;
Extension extensions<0..2^32-1>;
} ReInit;
]]>A member of the group applies a ReInit proposal by waiting for the committer to
send the Welcome message and by checking that the group_id and the parameters
of the new group corresponds to the ones specified in the proposal. The Welcome
message MUST specify exactly one pre-shared key with psktype = reinit, and with
psk_group_id and psk_epoch equal to the group_id and epoch of the
existing group after the Commit containing the reinit Proposal was processed.
The Welcome message may specify the inclusion of other pre-shared keys with a
psktype different from reinit.If a ReInit proposal is included in a Commit, it MUST be the only proposal
referenced by the Commit. If other non-ReInit proposals have been sent during
the epoch, the committer SHOULD prefer them over the ReInit proposal, allowing
the ReInit to be resent and applied in a subsequent epoch. The version field
in the ReInit proposal MUST be no less than the version for the current group.An ExternalInit proposal is used by new members that want to join a group by
using an external commit. This propsal can only be used in that context.;
} ExternalInit;
]]>A member of the group applies an ExternalInit message by initializing the next
epoch using an init secret computed as described in .
The kem_output field contains the required KEM output.Add and Remove proposals can be constructed and sent to the group by a party
that is outside the group. For example, a Delivery Service might propose to
remove a member of a group who has been inactive for a long time, or propose adding
a newly-hired staff member to a group representing a real-world team. Proposals
originating outside the group are identified by a preconfigured or
new_member SenderType in MLSPlaintext.ReInit proposals can also be sent to the group by a preconfigured sender, for
example to enforce a changed policy regarding MLS version or ciphersuite.The new_member SenderType is used for clients proposing that they themselves
be added. For this ID type the sender value MUST be zero and the Proposal type
MUST be Add. The MLSPlaintext MUST be signed with the private key corresponding
to the KeyPackage in the Add message. Recipients MUST verify that the
MLSPlaintext carrying the Proposal message is validly signed with this key.The preconfigured SenderType is reserved for signers that are pre-provisioned
to the clients within a group. If proposals with these sender IDs are to be
accepted within a group, the members of the group MUST be provisioned by the
application with a mapping between these IDs and authorized signing keys.
Recipients MUST verify that the MLSPlaintext carrying the Proposal message is
validly signed with the corresponding key. To ensure consistent handling of
external proposals, the application MUST ensure that the members of a group
have the same mapping and apply the same policies to external proposals.An external proposal MUST be sent as an MLSPlaintext
object, since the sender will not have the keys necessary to construct an
MLSCiphertext object.A Commit message initiates a new epoch for the group, based on a collection of
Proposals. It instructs group members to update their representation of the
state of the group by applying the proposals and advancing the key schedule.Each proposal covered by the Commit is included by a ProposalOrRef value, which
identifies the proposal to be applied by value or by reference. Proposals
supplied by value are included directly in the Commit object. Proposals
supplied by reference are specified by including the hash of the MLSPlaintext in
which the Proposal was sent, using the hash function from the group's
ciphersuite. For proposals supplied by value, the sender of the proposal is the
same as the sender of the Commit. Conversely, proposals sent by people other
than the committer MUST be included by reference.;
}
} ProposalOrRef;
struct {
ProposalOrRef proposals<0..2^32-1>;
optional path;
} Commit;
]]>A group member that has observed one or more proposals within an epoch MUST send
a Commit message before sending application data. This ensures, for example,
that any members whose removal was proposed during the epoch are actually
removed before any application data is transmitted.The sender of a Commit MUST include all valid proposals that it has received
during the current epoch. Invalid proposals include, for example, proposals with
an invalid signature or proposals that are semantically invalid, such as an Add
when the sender does not have the application-level permission to add new users.
If there are multiple proposals that apply to the same leaf, the committer
chooses one and includes only that one in the Commit, considering the rest
invalid. The committer MUST prefer any Remove received, or the most recent
Update for the leaf if there are no Removes. If there are multiple Add proposals
for the same client, the committer again chooses one to include and considers
the rest invalid.The Commit MUST NOT combine proposals sent within different epochs. In the event
that a valid proposal is omitted from the next Commit, the sender of the
proposal SHOULD retransmit it in the new epoch.A member of the group MAY send a Commit that references no proposals at all,
which would thus have an empty proposals vector. Such
a Commit resets the sender's leaf and the nodes along its direct path, and
provides forward secrecy and post-compromise security with regard to the sender
of the Commit. An Update proposal can be regarded as a "lazy" version of this
operation, where only the leaf changes and intermediate nodes are blanked out.The path field of a Commit message MUST be populated if the Commit covers at
least one Update or Remove proposal. The path field MUST also be populated
if the Commit covers no proposals at all (i.e., if the proposals vector
is empty). The path field MAY be omitted if the Commit covers only Add
proposals. In pseudocode, the logic for validating a Commit is as follows:To summarize, a Commit can have three different configurations, with different
uses:An "empty" Commit that references no proposals, which updates the committer's
contribution to the group and provides PCS with regard to the committer.An "add-only" Commit that references only Add proposals, in which the path is
optional. Such a commit provides PCS with regard to the committer only if
the path field is present.A "full" Commit that references proposals of any type, which provides FS with
regard to any removed members and PCS for the committer and any updated
members.A member of the group creates a Commit message and the corresponding Welcome
message at the same time, by taking the following steps:Construct an initial Commit object with the proposals
field populated from Proposals received during the current epoch, and empty
key_package and path fields.Generate a provisional GroupContext object by applying the proposals
referenced in the initial Commit object, as described in . Update
proposals are applied first, followed by Remove proposals, and then finally
Add proposals. Add proposals are applied in the order listed in the
proposals vector, and always to the leftmost unoccupied leaf in the tree, or
the right edge of the tree if all leaves are occupied. Note that the order in which different types of proposals are applied should
be updated by the implementation to include any new proposals added by
negotiated group extensions.PreSharedKey proposals are processed later when deriving the psk_secret for the Key
Schedule.Decide whether to populate the path field: If the path field is required
based on the proposals that are in the commit (see above), then it MUST be
populated. Otherwise, the sender MAY omit the path field at its discretion.If populating the path field: Create a UpdatePath using the new tree. Any
new member (from an add proposal) MUST be exluded from the resolution during
the computation of the UpdatePath. The GroupContext for this operation uses
the group_id, epoch, tree_hash, and confirmed_transcript_hash values
in the initial GroupContext object. Assign this UpdatePath to the path field in the Commit.Apply the UpdatePath to the tree, as described in
. Define commit_secret as the value
path_secret[n+1] derived from the path_secret[n] value assigned to
the root node.If not populating the path field: Set the path field in the Commit to the
null optional. Define commit_secret as the all-zero vector of the same
length as a path_secret value would be.Generate a new KeyPackage for the Committer's own leaf, with a
parent_hash extension. Store it in the ratchet tree and assign it to the
key_package field in the Commit object.If one or more PreSharedKey proposals are part of the commit, derive the psk_secret
as specified in , where the order of PSKs in the derivation
corresponds to the order of PreSharedKey proposals in the proposals vector.
Otherwise, set psk_secret to a zero-length octet string.Construct an MLSPlaintext object containing the Commit object. Sign the MLSPlaintext
using the current epoch's GroupContext as context. Use the signature, the
commit_secret and the psk_secret to advance the key schedule and compute
the confirmation_tag value in the MLSPlaintext.Update the tree in the provisional state by applying the direct pathConstruct a GroupInfo reflecting the new state:
Group ID, epoch, tree, confirmed transcript hash, and interim transcript
hash from the new stateThe confirmation_tag from the MLSPlaintext objectSign the GroupInfo using the member's private signing keyEncrypt the GroupInfo using the key and nonce derived from the joiner_secret
for the new epoch (see )For each new member in the group:
Identify the lowest common ancestor in the tree of the new member's
leaf node and the member sending the CommitIf the path field was populated above: Compute the path secret
corresponding to the common ancestor nodeCompute an EncryptedGroupSecrets object that encapsulates the init_secret
for the current epoch and the path secret (if present).Construct a Welcome message from the encrypted GroupInfo object, the encrypted
key packages, and any PSKs for which a proposal was included in the Commit. The
order of the psks MUST be the same as the order of PreSharedKey proposals in the
proposals vector.If a ReInit proposal was part of the Commit, the committer MUST create a new
group with the parameters specified in the ReInit proposal,
and with the same members as the original group.
The Welcome message MUST include a PreSharedKeyID with psktypereinit and with psk_group_id and psk_epoch corresponding to the current
group and the epoch after the commit was processed.A member of the group applies a Commit message by taking the following steps:Verify that the epoch field of the enclosing MLSPlaintext message is equal
to the epoch field of the current GroupContext objectVerify that the signature on the MLSPlaintext message verifies using the
public key from the credential stored at the leaf in the tree indicated by
the sender field.Verify that all PSKs specified in any PreSharedKey proposals in the proposals vector
are available.Generate a provisional GroupContext object by applying the proposals
referenced in the initial Commit object, as described in . Update
proposals are applied first, followed by Remove proposals, and then finally
Add proposals. Add proposals are applied in the order listed in the
proposals vector, and always to the leftmost unoccupied leaf in the tree, or
the right edge of the tree if all leaves are occupied. Note that the order in which different types of proposals are applied should
be updated by the implementation to include any new proposals added by
negotiated group extensions.Verify that the path value is populated if the proposals vector contains
any Update or Remove proposals, or if it's empty. Otherwise, the path value
MAY be omitted.If the path value is populated: Process the path value using the ratchet
tree the provisional GroupContext, to update the ratchet tree and generate the
commit_secret: Apply the UpdatePath to the tree, as described in
, and store key_package at the
Committer's leaf.Verify that the KeyPackage has a parent_hash extension and that its value
matches the new parent of the sender's leaf node.Define commit_secret as the value path_secret[n+1] derived from the
path_secret[n] value assigned to the root node.If the path value is not populated: Define commit_secret as the all-zero
vector of the same length as a path_secret value would be.Update the new GroupContext's confirmed and interim transcript hashes using the
new Commit.If the proposals vector contains any PreSharedKey proposals, derive the
psk_secret as specified in , where the order of PSKs in
the derivation corresponds to the order of PreSharedKey proposals in the
proposals vector. Otherwise, set psk_secret to 0.Use the commit_secret, the psk_secret, the provisional GroupContext, and
the init secret from the previous epoch to compute the epoch secret and
derived secrets for the new epoch.Use the confirmation_key for the new epoch to compute the confirmation tag
for this message, as described below, and verify that it is the same as the
confirmation_tag field in the MLSPlaintext object.If the above checks are successful, consider the updated GroupContext object
as the current state of the group.If the Commit included a ReInit proposal, the client MUST NOT use the group to
send messages anymore. Instead, it MUST wait for a Welcome message from the committer
and check that The version, cipher_suite and extensions fields of the new group
corresponds to the ones in the ReInit proposal, and that the version
is greater than or equal to that of the original group.The psks field in the Welcome message includes a PreSharedKeyID with
psktype = reinit, and psk_epoch and psk_group_id equal to the epoch
and group ID of the original group after processing the Commit.The confirmation tag value confirms that the members of the group have arrived
at the same state of the group:External Commits are a mechanism for new members (external parties that want to
become members of the group) to add themselves to a group, without requiring
that an existing member has to come online to issue a Commit that references an
Add Proposal.Whether existing members of the group will accept or reject an External Commit
follows the same rules that are applied to other handshake messages.New members can create and issue an External Commit if they have access to the
following information for the group's current epoch:group IDepoch IDciphersuitepublic tree hashinterim transcript hashgroup extensionsexternal public keyThis information is aggregated in a PublicGroupState object as follows:
struct {
CipherSuite cipher_suite;
opaque group_id<0..255>;
uint64 epoch;
opaque tree_hash<0..255>;
opaque interim_transcript_hash<0..255>;
Extension extensions<0..2^32-1>;
HPKEPublicKey external_pub;
} PublicGroupState;
Note that the tree_hash field is used the same way as in the Welcome message.
The full tree can be included via the ratchet_tree extension
.The information above are not deemed public data in general, but applications
can choose to make them available to new members in order to allow External
Commits.External Commits work like regular Commits, with a few differences:External Commits MUST reference an Add Proposal that adds the issuing new
member to the groupExternal Commits MUST contain a path field (and is therefore a "full"
Commit)External Commits MUST be signed by the new member. In particular, the
signature on the enclosing MLSPlaintext MUST verify using the public key for
the credential in the leaf_key_package of the path field.An external commit MUST reference no more than one ExternalInit proposal, and the
ExternalInit proposal MUST be supplied by value, not by reference. When
processing a Commit, both existing and new members MUST use the external init
secret as described in .The sender type for the MLSPlaintext encapsulating the External Commit MUST be
new_memberIf the Add Proposal is also issued by the new member, its member SenderType
MUST be new_memberThe sender of a Commit message is responsible for sending a Welcome message to
any new members added via Add proposals. The Welcome message provides the new
members with the current state of the group, after the application of the Commit
message. The new members will not be able to decrypt or verify the Commit
message, but will have the secrets they need to participate in the epoch
initiated by the Commit message.In order to allow the same Welcome message to be sent to all new members,
information describing the group is encrypted with a symmetric key and nonce
randomly chosen by the sender. This key and nonce are then encrypted to each
new member using HPKE. In the same encrypted package, the committer transmits
the path secret for the lowest node contained in the direct paths of both the
committer and the new member. This allows the new member to compute private
keys for nodes in its direct path that are being reset by the corresponding
Commit.If the sender of the Welcome message wants the receiving member to include a PSK
in the derivation of the epoch_secret, they can populate the psks field indicating which
PSK to use.;
uint64 epoch;
opaque tree_hash<0..255>;
opaque confirmed_transcript_hash<0..255>;
Extension extensions<0..2^32-1>;
MAC confirmation_tag;
uint32 signer_index;
opaque signature<0..2^16-1>;
} GroupInfo;
struct {
opaque path_secret<1..255>;
} PathSecret;
struct {
opaque joiner_secret<1..255>;
optional path_secret;
optional psks;
} GroupSecrets;
struct {
opaque key_package_hash<1..255>;
HPKECiphertext encrypted_group_secrets;
} EncryptedGroupSecrets;
struct {
ProtocolVersion version = mls10;
CipherSuite cipher_suite;
EncryptedGroupSecrets secrets<0..2^32-1>;
opaque encrypted_group_info<1..2^32-1>;
} Welcome;
]]>The client processing a Welcome message will need to have a copy of the group's
ratchet tree. The tree can be provided in the Welcome message, in an extension
of type ratchet_tree. If it is sent otherwise (e.g., provided by a caching
service on the Delivery Service), then the client MUST download the tree before
processing the Welcome.On receiving a Welcome message, a client processes it using the following steps:Identify an entry in the secrets array where the key_package_hash
value corresponds to one of this client's KeyPackages, using the hash
indicated by the cipher_suite field. If no such field exists, or if the
ciphersuite indicated in the KeyPackage does not match the one in the
Welcome message, return an error.Decrypt the encrypted_group_secrets using HPKE with the algorithms indicated
by the ciphersuite and the HPKE private key corresponding to the GroupSecrets.
If a PreSharedKeyID is part of the GroupSecrets and the client is not in
possession of the corresponding PSK, return an error.From the joiner_secret in the decrypted GroupSecrets object and the PSKs
specified in the GroupSecrets, derive the member_secret and using that the
welcome_secret, welcome_key, and welcome_nonce. Use the key and nonce to
decrypt the encrypted_group_info field.Verify the signature on the GroupInfo object. The signature input comprises
all of the fields in the GroupInfo object except the signature field. The
public key and algorithm are taken from the credential in the leaf node at
position signer_index. If this verification fails, return an error.Verify the integrity of the ratchet tree. Verify that the tree hash of the ratchet tree matches the tree_hash field
in the GroupInfo.For each non-empty parent node, verify that exactly one of the node's
children are non-empty and have the hash of this node set as their
parent_hash value (if the child is another parent) or has a parent_hash
extension in the KeyPackage containing the same value (if the child is a
leaf).For each non-empty leaf node, verify the signature on the KeyPackage.Identify a leaf in the tree array (any even-numbered node) whose
key_package field is identical to the the KeyPackage. If no such field
exists, return an error. Let index represent the index of this node among
the leaves in the tree, namely the index of the node in the tree array
divided by two.Construct a new group state using the information in the GroupInfo object.
The new member's position in the tree is index, as defined above. In
particular, the confirmed transcript hash for the new state is the
prior_confirmed_transcript_hash in the GroupInfo object. Update the leaf at index index with the private key corresponding to the
public key in the node.If the path_secret value is set in the GroupSecrets object: Identify the
lowest common ancestor of the leaves at index and at
GroupInfo.signer_index. Set the private key for this node to the
private key derived from the path_secret.For each parent of the common ancestor, up to the root of the tree, derive
a new path secret and set the private key for the node to the private key
derived from the path secret. The private key MUST be the private key
that corresponds to the public key in the node.Use the joiner_secret from the GroupSecrets object to generate the epoch secret
and other derived secrets for the current epoch.Set the confirmed transcript hash in the new state to the value of the
confirmed_transcript_hash in the GroupInfo.Verify the confirmation tag in the GroupInfo using the derived confirmation
key and the confirmed_transcript_hash from the GroupInfo.Use the confirmed transcript hash and confirmation tag to compute the interim
transcript hash in the new state.By default, a GroupInfo message only provides the joiner with a commitment
to the group's ratchet tree. In order to process or generate handshake
messages, the joiner will need to get a copy of the ratchet tree from some other
source. (For example, the DS might provide a cached copy.) The inclusion of
the tree hash in the GroupInfo message means that the source of the ratchet
tree need not be trusted to maintain the integrity of tree.In cases where the application does not wish to provide such an external source,
the whole public state of the ratchet tree can be provided in an extension of
type ratchet_tree, containing a ratchet_tree object of the following form: ratchet_tree<1..2^32-1>;
]]>The presence of a ratchet_tree extension in a GroupInfo message does not
result in any changes to the GroupContext extensions for the group. The ratchet
tree provided is simply stored by the client and used for MLS operations.If this extension is not provided in a Welcome message, then the client will
need to fetch the ratchet tree over some other channel before it can generate or
process Commit messages. Applications should ensure that this out-of-band
channel is provided with security protections equivalent to the protections that
are afforded to Proposal and Commit messages. For example, an application that
encrypts Proposal and Commit messages might distribute ratchet trees encrypted
using a key exchanged over the MLS channel.This protocol includes a mechanism for negotiating extension parameters similar
to the one in TLS . In TLS, extension negotiation is one-to-one: The
client offers extensions in its ClientHello message, and the server expresses
its choices for the session with extensions in its ServerHello and
EncryptedExtensions messages. In MLS, extensions appear in the following
places:In KeyPackages, to describe client capabilities and aspects of their
participation in the group (once in the ratchet tree)In the Welcome message, to tell new members of a group what parameters are
being used by the groupIn the GroupContext object, to ensure that all members of the group have the
same view of the parameters in useIn other words, clients advertise their capabilities in KeyPackage
extensions, the creator of the group expresses its choices for the group in
Welcome extensions, and the GroupContext confirms that all members of the group
have the same view of the group's extensions.This extension mechanism is designed to allow for secure and forward-compatible
negotiation of extensions. For this to work, implementations MUST correctly
handle extensible fields:A client that posts a KeyPackage MUST support all parameters advertised in
it. Otherwise, another client might fail to interoperate by selecting one of
those parameters.A client initiating a group MUST ignore all unrecognized ciphersuites,
extensions, and other parameters. Otherwise, it may fail to interoperate with
newer clients.A client adding a new member to a group MUST verify that the KeyPackage
for the new member contains extensions that are consistent with the group's
extensions. For each extension in the GroupContext, the KeyPackage MUST
have an extension of the same type, and the contents of the extension MUST be
consistent with the value of the extension in the GroupContext, according to
the semantics of the specific extension.If any extension in a GroupInfo message is unrecognized (i.e., not contained
in the corresponding KeyPackage), then the client MUST reject the Welcome
message and not join the group.The extensions populated into a GroupContext object are drawn from those in
the GroupInfo object, according to the definitions of those extensions.Note that the latter two requirements mean that all MLS extensions are
mandatory, in the sense that an extension in use by the group MUST be supported
by all members of the group.This document does not define any way for the parameters of the group to change
once it has been created; such a behavior could be implemented as an extension.Each Commit message is premised on a given starting state,
indicated by the epoch field of the enclosing MLSPlaintext
message. If the changes implied by a Commit messages are made
starting from a different state, the results will be incorrect.This need for sequencing is not a problem as long as each time a
group member sends a Commit message, it is based on the most
current state of the group. In practice, however, there is a risk
that two members will generate Commit messages simultaneously,
based on the same state.When this happens, there is a need for the members of the group to
deconflict the simultaneous Commit messages. There are two
general approaches:Have the Delivery Service enforce a total orderHave a signal in the message that clients can use to break tiesAs long as Commit messages cannot be merged, there is a risk of
starvation. In a sufficiently busy group, a given member may never
be able to send a Commit message, because he always loses to other
members. The degree to which this is a practical problem will depend
on the dynamics of the application.It might be possible, because of the non-contributivity of intermediate
nodes, that Commit messages could be applied one after the other
without the Delivery Service having to reject any Commit message,
which would make MLS more resilient regarding the concurrency of
Commit messages.
The Messaging system can decide to choose the order for applying
the state changes. Note that there are certain cases (if no total
ordering is applied by the Delivery Service) where the ordering is
important for security, ie. all updates must be executed before
removes.Regardless of how messages are kept in sequence, implementations
MUST only update their cryptographic state when valid Commit
messages are received.
Generation of Commit messages MUST NOT modify a client's state, since the
endpoint doesn't know at that time whether the changes implied by
the Commit message will succeed or not.With this approach, the Delivery Service ensures that incoming
messages are added to an ordered queue and outgoing messages are
dispatched in the same order. The server is trusted to break ties
when two members send a Commit message at the same time.Messages should have a counter field sent in clear-text that can
be checked by the server and used for tie-breaking. The counter
starts at 0 and is incremented for every new incoming message.
If two group members send a message with the same counter, the
first message to arrive will be accepted by the server and the
second one will be rejected. The rejected message needs to be sent
again with the correct counter number.To prevent counter manipulation by the server, the counter's
integrity can be ensured by including the counter in a signed
message envelope.This applies to all messages, not only state changing messages.Order enforcement can be implemented on the client as well,
one way to achieve it is to use a two step update protocol: the
first client sends a proposal to update and the proposal is
accepted when it gets 50%+ approval from the rest of the group,
then it sends the approved update. Clients which didn't get
their proposal accepted, will wait for the winner to send their
update before retrying new proposals.While this seems safer as it doesn't rely on the server, it is
more complex and harder to implement. It also could cause starvation
for some clients if they keep failing to get their proposal accepted.The primary purpose of the Handshake protocol is to provide an
authenticated group key exchange to clients. In order to protect
Application messages sent among the members of a group, the Application
secret provided by the Handshake key schedule is used to derive nonces
and encryption keys for the Message Protection Layer according to
the Application Key Schedule. That is, each epoch is equipped with
a fresh Application Key Schedule which consist of a tree of Application
Secrets as well as one symmetric ratchet per group member.Each client maintains their own local copy of the Application Key
Schedule for each epoch during which they are a group member. They
derive new keys, nonces and secrets as needed while deleting old
ones as soon as they have been used.Application messages MUST be protected with the Authenticated-Encryption
with Associated-Data (AEAD) encryption scheme associated with the
MLS ciphersuite using the common framing mechanism.
Note that "Authenticated" in this context does not mean messages are
known to be sent by a specific client but only from a legitimate
member of the group.
To authenticate a message from a particular member, signatures are
required. Handshake messages MUST use asymmetric signatures to strongly
authenticate the sender of a message.The group members MUST use the AEAD algorithm associated with
the negotiated MLS ciphersuite to AEAD encrypt and decrypt their
Application messages according to the Message Framing section.The group identifier and epoch allow a recipient to know which group secrets
should be used and from which Epoch secret to start computing other secrets
and keys. The sender identifier is used to identify the member's
symmetric ratchet from the initial group Application secret. The application
generation field is used to determine how far into the ratchet to iterate in
order to reproduce the required AEAD keys and nonce for performing decryption.Application messages SHOULD be padded to provide some resistance
against traffic analysis techniques over encrypted traffic.
While MLS might deliver the same payload less frequently across
a lot of ciphertexts than traditional web servers, it might still provide
the attacker enough information to mount an attack. If Alice asks Bob:
"When are we going to the movie ?" the answer "Wednesday" might be leaked
to an adversary by the ciphertext length. An attacker expecting Alice to
answer Bob with a day of the week might find out the plaintext by
correlation between the question and the length.Similarly to TLS 1.3, if padding is used, the MLS messages MUST be
padded with zero-valued bytes before AEAD encryption. Upon AEAD decryption,
the length field of the plaintext is used to compute the number of bytes
to be removed from the plaintext to get the correct data.
As the padding mechanism is used to improve protection against traffic
analysis, removal of the padding SHOULD be implemented in a "constant-time"
manner at the MLS layer and above layers to prevent timing side-channels that
would provide attackers with information on the size of the plaintext.
The padding length length_of_padding can be chosen at the time of the message
encryption by the sender. Recipients can calculate the padding size from knowing
the total size of the ApplicationPlaintext and the length of the content.During each epoch senders MUST NOT encrypt more data than permitted by the
security bounds of the AEAD scheme used.Note that each change to the Group through a Handshake message will also set a
new encryption_secret. Hence this change MUST be applied before encrypting
any new application message. This is required both to ensure that any users
removed from the group can no longer receive messages and to (potentially)
recover confidentiality and authenticity for future messages despite a past
state compromise.Since each Application message contains the group identifier, the epoch and a
message counter, a client can receive messages out of order.
If they are able to retrieve or recompute the correct AEAD decryption key
from currently stored cryptographic material clients can decrypt
these messages.For usability, MLS clients might be required to keep the AEAD key
and nonce for a certain amount of time to retain the ability to decrypt
delayed or out of order messages, possibly still in transit while a
decryption is being done.The security goals of MLS are described in .
We describe here how the protocol achieves its goals at a high level,
though a complete security analysis is outside of the scope of this
document.Group secrets are partly derived from the output of a ratchet tree. Ratchet
trees work by assigning each member of the group to a leaf in the tree and
maintaining the following property: the private key of a node in the tree is
known only to members of the group that are assigned a leaf in the node's
subtree. This is called the ratchet tree invariant and it makes it possible to
encrypt to all group members except one, with a number of ciphertexts that's
logarithmic in the number of group members.The ability to efficiently encrypt to all members except one allows members to
be securely removed from a group. It also allows a member to rotate their
keypair such that the old private key can no longer be used to decrypt new
messages.The first form of authentication we provide is that group members can verify a
message originated from one of the members of the group. For encrypted messages,
this is guaranteed because messages are encrypted with an AEAD under a key
derived from the group secrets. For plaintext messages, this is guaranteed by
the use of a membership_tag which constitutes a MAC over the message, under a
key derived from the group secrets.The second form of authentication is that group members can verify a message
originated from a particular member of the group. This is guaranteed by a
digital signature on each message from the sender's identity key.Post-compromise security is provided between epochs by members regularly
updating their leaf key in the ratchet tree. Updating their leaf key prevents
group secrets from continuing to be encrypted to previously compromised public
keys.Forward-secrecy between epochs is provided by deleting private keys from past
version of the ratchet tree, as this prevents old group secrets from being
re-derived. Forward secrecy within an epoch is provided by deleting message
encryption keys once they've been used to encrypt or decrypt a message.Post-compromise security is also provided for new groups by members regularly
generating new InitKeys and uploading them to the Delivery Service, such that
compromised key material won't be used when the member is added to a new group.InitKeys are intended to be used only once. That is, once an InitKey has been
used to introduce the corresponding client to a group, it SHOULD be deleted from
the InitKey publication system. Reuse of InitKeys can lead to replay attacks.An application MAY allow for reuse of a "last resort" InitKey in order to
prevent denial of service attacks. Since an InitKey is needed to add a client
to a new group, an attacker could prevent a client being added to new groups by
exhausting all available InitKeys.This document requests the creation of the following new IANA registries:MLS Ciphersuites ()MLS Extension Types ()MLS Credential Types ()All of these registries should be under a heading of "Messaging Layer Security",
and assignments are made via the Specification Required policy . See
for additional information about the MLS Designated Experts (DEs).RFC EDITOR: Please replace XXXX throughout with the RFC number assigned to
this documentA ciphersuite is a combination of a protocol version and the set of
cryptographic algorithms that should be used.Ciphersuite names follow the naming convention:Where VALUE is represented as a sixteen-bit integer:ComponentContentsMLSThe string "MLS" followed by the major and minor version, e.g. "MLS10"LVLThe security levelKEMThe KEM algorithm used for HPKE in TreeKEM group operationsAEADThe AEAD algorithm used for HPKE and message protectionHASHThe hash algorithm used for HPKE and the MLS transcript hashSIGThe Signature algorithm used for message authenticationThe columns in the registry are as follows:Value: The numeric value of the ciphersuiteName: The name of the ciphersuiteRecommended: Whether support for this extension is recommended by the IETF MLS
WG. Valid values are "Y" and "N". The "Recommended" column is assigned a
value of "N" unless explicitly requested, and adding a value with a
"Recommended" value of "Y" requires Standards Action . IESG Approval
is REQUIRED for a Y->N transition.Reference: The document where this ciphersuite is definedInitial contents:ValueNameRecommendedReference0x0000RESERVEDN/ARFC XXXX0x0001MLS10_128_DHKEMX25519_AES128GCM_SHA256_Ed25519YRFC XXXX0x0002MLS10_128_DHKEMP256_AES128GCM_SHA256_P256YRFC XXXX0x0003MLS10_128_DHKEMX25519_CHACHA20POLY1305_SHA256_Ed25519YRFC XXXX0x0004MLS10_256_DHKEMX448_AES256GCM_SHA512_Ed448YRFC XXXX0x0005MLS10_256_DHKEMP521_AES256GCM_SHA512_P521YRFC XXXX0x0006MLS10_256_DHKEMX448_CHACHA20POLY1305_SHA512_Ed448YRFC XXXX0xff00 - 0xffffReserved for Private UseN/ARFC XXXXAll of these ciphersuites use HMAC as their MAC function, with
different hashes per ciphersuite. The mapping of ciphersuites to HPKE
primitives, HMAC hash functions, and TLS signature schemes is as follows
:ValueKEMKDFAEADHashSignature0x00010x00200x00010x0001SHA256ed255190x00020x00100x00010x0001SHA256ecdsa_secp256r1_sha2560x00030x00200x00010x0003SHA256ed255190x00040x00210x00030x0002SHA512ed4480x00050x00120x00030x0002SHA512ecdsa_secp521r1_sha5120x00060x00210x00030x0003SHA512ed448The hash used for the MLS transcript hash is the one referenced in the
ciphersuite name. In the ciphersuites defined above, "SHA256" and "SHA512"
refer to the SHA-256 and SHA-512 functions defined in .It is advisable to keep the number of ciphersuites low to increase the chances
clients can interoperate in a federated environment, therefore the ciphersuites
only inlcude modern, yet well-established algorithms. Depending on their
requirements, clients can choose between two security levels (roughly 128-bit
and 256-bit). Within the security levels clients can choose between faster
X25519/X448 curves and FIPS 140-2 compliant curves for Diffie-Hellman key
negotiations. Additionally clients that run predominantly on mobile processors
can choose ChaCha20Poly1305 over AES-GCM for performance reasons. Since
ChaCha20Poly1305 is not listed by FIPS 140-2 it is not paired with FIPS 140-2
compliant curves. The security level of symmetric encryption algorithms and hash
functions is paired with the security level of the curves.The mandatory-to-implement ciphersuite for MLS 1.0 is
MLS10\_128\_HPKE25519\_AES128GCM\_SHA256\_Ed25519 which uses
Curve25519 for key exchange, AES-128-GCM for HPKE, HKDF over SHA2-256,
AES for metadata masking, and Ed25519 for signatures.Values with the first byte 255 (decimal) are reserved for Private Use.New ciphersuite values are assigned by IANA as described in
.This registry lists identifiers for extensions to the MLS protocol. The
extension type field is two bytes wide, so valid extension type values are in
the range 0x0000 to 0xffff.Template:Value: The numeric value of the extension typeName: The name of the extension typeMessage(s): The messages in which the extension may appear, drawn from the following
list: KP: KeyPackage messagesGI: GroupInfo objectsRecommended: Whether support for this extension is recommended by the IETF MLS
WG. Valid values are "Y" and "N". The "Recommended" column is assigned a
value of "N" unless explicitly requested, and adding a value with a
"Recommended" value of "Y" requires Standards Action . IESG Approval
is REQUIRED for a Y->N transition.Reference: The document where this extension is definedInitial contents:ValueNameMessage(s)RecommendedReference0x0000RESERVEDN/AN/ARFC XXXX0x0001capabilitiesKPYRFC XXXX0x0002lifetimeKPYRFC XXXX0x0003key_idKPYRFC XXXX0x0004parent_hashKPYRFC XXXX0x0005ratchet_treeGIYRFC XXXX0xff00 - 0xffffReserved for Private UseN/AN/ARFC XXXXThis registry lists identifiers for types of credentials that can be used for
authentication in the MLS protocol. The extension type field is two bytes wide,
so valid extension type values are in the range 0x0000 to 0xffff.Template:Value: The numeric value of the credential typeName: The name of the credential typeRecommended: Whether support for this extension is recommended by the IETF MLS
WG. Valid values are "Y" and "N". The "Recommended" column is assigned a
value of "N" unless explicitly requested, and adding a value with a
"Recommended" value of "Y" requires Standards Action . IESG Approval
is REQUIRED for a Y->N transition.Reference: The document where this extension is definedInitial contents:ValueNameRecommendedReference0x0000RESERVEDN/ARFC XXXX0x0001basicYRFC XXXX0x0002x509YRFC XXXX0xff00 - 0xffffReserved for Private UseN/ARFC XXXXSpecification Required registry requests are registered
after a three-week review period on the MLS DEs' mailing list:
mls-reg-review@ietf.org, on the advice of one or more of the MLS DEs. However,
to allow for the allocation of values prior to publication, the MLS
DEs may approve registration once they are satisfied that such a
specification will be published.Registration requests sent to the MLS DEs mailing list for review
SHOULD use an appropriate subject (e.g., "Request to register value
in MLS Bar registry").Within the review period, the MLS DEs will either approve or deny
the registration request, communicating this decision to the MLS DEs
mailing list and IANA. Denials SHOULD include an explanation and, if
applicable, suggestions as to how to make the request successful.
Registration requests that are undetermined for a period longer than
21 days can be brought to the IESG's attention for resolution using
the iesg@ietf.org mailing list.Criteria that SHOULD be applied by the MLS DEs includes determining
whether the proposed registration duplicates existing functionality,
whether it is likely to be of general applicability or useful only
for a single application, and whether the registration description
is clear. For example, the MLS DEs will apply the ciphersuite-related
advisory found in .IANA MUST only accept registry updates from the MLS DEs and SHOULD
direct all requests for registration to the MLS DEs' mailing list.It is suggested that multiple MLS DEs be appointed who are able to
represent the perspectives of different applications using this
specification, in order to enable broadly informed review of
registration decisions. In cases where a registration decision could
be perceived as creating a conflict of interest for a particular
MLS DE, that MLS DE SHOULD defer to the judgment of the other MLS DEs.Joel Alwen
Wickr
joel.alwen@wickr.comKarthikeyan Bhargavan
INRIA
karthikeyan.bhargavan@inria.frCas Cremers
University of Oxford
cas.cremers@cs.ox.ac.ukAlan Duric
Wire
alan@wire.comBritta Hale
Naval Postgraduate School
britta.hale@nps.eduSrinivas Inguva
Twitter
singuva@twitter.comKonrad Kohbrok
Aalto University
konrad.kohbrok@datashrine.deAlbert Kwon
MIT
kwonal@mit.eduBrendan McMillion
Cloudflare
brendan@cloudflare.comEric Rescorla
Mozilla
ekr@rtfm.comMichael Rosenberg
Trail of Bits
michael.rosenberg@trailofbits.comThyla van der Merwe
Royal Holloway, University of London
thyla.van.der@merwe.techKey words for use in RFCs to Indicate Requirement LevelsIn 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.Ambiguity of Uppercase vs Lowercase in RFC 2119 Key WordsRFC 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.Hybrid Public Key EncryptionThis document describes a scheme for hybrid public-key encryption (HPKE). This scheme provides authenticated public key encryption of arbitrary-sized plaintexts for a recipient public key. HPKE works for any combination of an asymmetric key encapsulation mechanism (KEM), key derivation function (KDF), and authenticated encryption with additional data (AEAD) encryption function. We provide instantiations of the scheme using widely-used and efficient primitives, such as Elliptic Curve Diffie-Hellman key agreement, HKDF, and SHA2.The Transport Layer Security (TLS) Protocol Version 1.3This document specifies version 1.3 of the Transport Layer Security (TLS) protocol. TLS allows client/server applications to communicate over the Internet in a way that is designed to prevent eavesdropping, tampering, and message forgery.This document updates RFCs 5705 and 6066, and obsoletes RFCs 5077, 5246, and 6961. This document also specifies new requirements for TLS 1.2 implementations.Guidelines for Writing an IANA Considerations Section in RFCsMany protocols make use of points of extensibility that use constants to identify various protocol parameters. To ensure that the values in these fields do not have conflicting uses and to promote interoperability, their allocations are often coordinated by a central record keeper. For IETF protocols, that role is filled by the Internet Assigned Numbers Authority (IANA).To make assignments in a given registry prudently, guidance describing the conditions under which new values should be assigned, as well as when and how modifications to existing values can be made, is needed. This document defines a framework for the documentation of these guidelines by specification authors, in order to assure that the provided guidance for the IANA Considerations is clear and addresses the various issues that are likely in the operation of a registry.This is the third edition of this document; it obsoletes RFC 5226.HMAC: Keyed-Hashing for Message AuthenticationThis document describes HMAC, a mechanism for message authentication using cryptographic hash functions. HMAC can be used with any iterative cryptographic hash function, e.g., MD5, SHA-1, in combination with a secret shared key. The cryptographic strength of HMAC depends on the properties of the underlying hash function. This memo provides information for the Internet community. This memo does not specify an Internet standard of any kindOn Ends-to-Ends Encryption: Asynchronous Group Messaging with Strong Security GuaranteesA Formal Security Analysis of the Signal Messaging ProtocolThe Double Ratchet AlgorithmElliptic Curve Cryptography, Standards for Efficient Cryptography Group, ver. 2Secure Hash StandardThe Messaging Layer Security (MLS) ArchitectureThis document describes the reference architecture, functional and security requirements for the Messaging Layer Security (MLS) protocol. MLS provides a security layer for group messaging applications, where the number of clients ranges from two to many. It is meant to protect against eavesdropping, tampering, and message forgery. Discussion Venues This note is to be removed before publishing as an RFC. Source for this draft and an issue tracker can be found at https://github.com/mlswg/mls-architecture.Certificate Transparency Version 2.0This document describes version 2.0 of the Certificate Transparency (CT) protocol for publicly logging the existence of Transport Layer Security (TLS) server certificates as they are issued or observed, in a manner that allows anyone to audit certification authority (CA) activity and notice the issuance of suspect certificates as well as to audit the certificate logs themselves. The intent is that eventually clients would refuse to honor certificates that do not appear in a log, effectively forcing CAs to add all issued certificates to the logs. This document obsoletes RFC 6962. It also specifies a new TLS extension that is used to send various CT log artifacts. Logs are network services that implement the protocol operations for submissions and queries that are defined in this document.Edwards-Curve Digital Signature Algorithm (EdDSA)This document describes elliptic curve signature scheme Edwards-curve Digital Signature Algorithm (EdDSA). The algorithm is instantiated with recommended parameters for the edwards25519 and edwards448 curves. An example implementation and test vectors are provided.I Know Why You Went to the Clinic: Risks and Realization of HTTPS Traffic AnalysisHTTPS traffic analysis and client identification using passive SSL/TLS fingerprintingOne benefit of using left-balanced trees is that they admit a simple
flat array representation. In this representation, leaf nodes are
even-numbered nodes, with the n-th leaf at 2*n. Intermediate nodes
are held in odd-numbered nodes. For example, an 11-element tree has
the following structure:This allows us to compute relationships between tree nodes simply by
manipulating indices, rather than having to maintain complicated
structures in memory, even for partial trees. The basic
rule is that the high-order bits of parent and child nodes have the
following relation (where x is an arbitrary bit string): left=00x, right=10x
]]>The following python code demonstrates the tree computations
necessary for MLS. Test vectors can be derived from the diagram
above.> k) > 0:
k += 1
return k-1
# The level of a node in the tree. Leaves are level 0, their parents are
# level 1, etc. If a node's children are at different levels, then its
# level is the max level of its children plus one.
def level(x):
if x & 0x01 == 0:
return 0
k = 0
while ((x >> k) & 0x01) == 1:
k += 1
return k
# The number of nodes needed to represent a tree with n leaves.
def node_width(n):
if n == 0:
return 0
else:
return 2*(n - 1) + 1
# The index of the root node of a tree with n leaves.
def root(n):
w = node_width(n)
return (1 << log2(w)) - 1
# The left child of an intermediate node. Note that because the tree is
# left-balanced, there is no dependency on the size of the tree.
def left(x):
k = level(x)
if k == 0:
raise Exception('leaf node has no children')
return x ^ (0x01 << (k - 1))
# The right child of an intermediate node. Depends on the number of
# leaves because the straightforward calculation can take you beyond the
# edge of the tree.
def right(x, n):
k = level(x)
if k == 0:
raise Exception('leaf node has no children')
r = x ^ (0x03 << (k - 1))
while r >= node_width(n):
r = left(r)
return r
# The immediate parent of a node. May be beyond the right edge of the
# tree.
def parent_step(x):
k = level(x)
b = (x >> (k + 1)) & 0x01
return (x | (1 << k)) ^ (b << (k + 1))
# The parent of a node. As with the right child calculation, we have to
# walk back until the parent is within the range of the tree.
def parent(x, n):
if x == root(n):
raise Exception('root node has no parent')
p = parent_step(x)
while p >= node_width(n):
p = parent_step(p)
return p
# The other child of the node's parent.
def sibling(x, n):
p = parent(x, n)
if x < p:
return right(p, n)
else:
return left(p)
# The direct path of a node, ordered from leaf to root.
def direct_path(x, n):
r = root(n)
if x == r:
return []
d = []
while x != r:
x = parent(x, n)
d.append(x)
return d
# The copath of a node, ordered from leaf to root.
def copath(x, n):
if x == root(n):
return []
d = direct_path(x, n)
d.insert(0, x)
d.pop()
return [sibling(y, n) for y in d]
# The common ancestor of two nodes is the lowest node that is in the
# direct paths of both leaves.
def common_ancestor_semantic(x, y, n):
dx = set([x]) | set(direct_path(x, n))
dy = set([y]) | set(direct_path(y, n))
dxy = dx & dy
if len(dxy) == 0:
raise Exception('failed to find common ancestor')
return min(dxy, key=level)
# The common ancestor of two nodes is the lowest node that is in the
# direct paths of both leaves.
def common_ancestor_direct(x, y, _):
# Handle cases where one is an ancestor of the other
lx, ly = level(x)+1, level(y)+1
if (lx <= ly) and (x>>ly == y>>ly):
return y
elif (ly <= lx) and (x>>lx == y>>lx):
return x
# Handle other cases
xn, yn = x, y
k = 0
while xn != yn:
xn, yn = xn >> 1, yn >> 1
k += 1
return (xn << k) + (1 << (k-1)) - 1
]]>