independent

roland@catalogix.se

Microsoft

mbj@microsoft.com https://self-issued.info/

Sikt

Andreas.Solberg@sikt.no https://www.linkedin.com/in/andreassolberg/

Yubico

ve7jtb@ve7jtb.com http://www.thread-safe.com/

independent

demarcog83@gmail.com https://www.linkedin.com/in/giuseppe-de-marco-bb054245/

Connect2id

vladimir@connect2id.com https://twitter.com/dzhuvi

OpenID Connect Working Group OpenID Connect Federation A federation can be expressed as an agreement between parties that trust each other. In bilateral federations, direct trust can be established between two organizations belonging to the same federation. In a multilateral federation, bilateral agreements might not be practical, in which case, trust can be mediated by a third party. That is the model used in this specification. An Entity in the federation must be able to trust that other entities it is interacting with belong to the same federation. It must also be able to trust that the information the other entities publish about themselves has not been tampered with during transport and that it adheres to the federation's policies. This specification describes the basic components to build a multilateral federation and it provides a guide on how to apply them when the underlying protocol used is OpenID Connect.

This specification describes how two entities that would like to interact can dynamically fetch and resolve trust and metadata for a given protocol through the use of third-party Trust Anchor. A trust anchor is an Entity whose main purpose is to issue statements about entities, such as OpenID Connect Relying Parties, OpenID Providers, and participating organizations. An identity federation can be realized using this specification using one or more levels of trust issuers. This specification does not mandate a specific way or restrict how a federation may be built. Instead, the specification provides the basic technical trust infrastructure building blocks needed to build a dynamic and distributed trust network such as a federation. Note that this specification only concerns itself with how entities in a federation get to know about each other. Furthermore, note that a company, as with any real-world organization, MAY be represented by more than one Entity in a federation. It is also true that an Entity can be part of more than one federation. OpenID Connect Federation Trust Chains rely on cryptographically signed JSON Web Token (JWT) documents, and the Trust Chain does not at all rely on TLS to establish trust.

The key words "MUST", "MUST NOT", "REQUIRED", "SHALL", "SHALL NOT", "SHOULD", "SHOULD NOT", "RECOMMENDED", "MAY", and "OPTIONAL" in this document are to be interpreted as described in .

This specification uses the terms "Claim Name", "Claim Value", "JSON Web Token (JWT)", defined by JSON Web Token (JWT) and the terms "OpenID Provider (OP)" and "Relying Party (RP)" defined by OpenID Connect Core 1.0. This specification also defines the following terms: Something that has a separate and distinct existence and that can be identified in a context. All entities in an OpenID Connect federation MUST have a globally unique identifier. A URI that is globally unique and that is bound to one Entity. Signed JWT that contains the information needed for an Entity to participate in federation(s), including metadata about itself and policies that apply to other entities that it is authoritative for. An Entity Statement issued by an Entity about itself. It contains the entities' signing keys and further data used to control the Trust Chain resolution process, such as authority hints. An organization that is authoritative for a federation. A federation operator administers the Trust Anchor(s) for entities in its federation. An Entity that issues an Entity Statement that appears somewhere in between those issued by the Trust Anchor and the Leaf Entity in a Trust Chain. An Entity defined by a certain protocol, e.g., OpenID Connect Relying Party or Provider. An Entity that represents a trusted third party. A process that starts with an Entity Identifier and collects a number of Entity Statements until the chosen Trust Anchor is reached. From the collected Entity Statements a Trust Chain is constructed and verified. The end result of the Federation Entity Discovery is that the Leaf Entity's metadata is constructed from the Trust Chain. A sequence of Entity Statements that represents a chain starting at a Leaf Entity and ending in a Trust Anchor. A trust mark represents a statement of conformance to a well-scoped set of trust and/or interoperability requirements as determined by an accreditation authority. Keys that an Entity uses to sign its Entity Configuration. These keys are also used by an entity to sign Entity Statements.

The basic component is the Entity Statement, which is a cryptographically signed JSON Web Token (JWT). A set of Entity Statements can form a path from a Leaf Entity to a Trust Anchor. Entity Configurations issued by Entities about themselves control the Trust Chain resolution process. The Entity Configuration of a Leaf Entity contains one or more references to its superiors, found in the authority_hints parameter. These references can be used to download the Entity Configuration of each superior. One or more Entity Configurations can occur, during the Federation Entity Discovery, until the Trust Anchor is reached. The Trust Anchor and its intermediaries issue Entity Statements about their descendants. The sequence of Entity Configurations and Entity Statements that validate the relationship between a superior and a descendant, along a path towards the Trust Anchor, forms the proof that a Leaf Entity is part of the Federation configured by the Trust Anchor. The chain that links the statements to one another is verifiable with the signature of each statement, as described in . With a verified Trust Chain, the federation policy is finally applied and the metadata describing a Leaf Entity within the Federation is then derived. How this is done is described in . This specification deals with trust operations; it doesn't cover or touch protocol operations outside those of metadata derivation and exchange. In OpenID Connect terms, these are protocol operations other than OpenID Connect Discovery 1.0 and OpenID Connect Dynamic Client Registration 1.0 . OpenID Connect is used in all of the examples in this specification, however this doesn't mean that this specification can only be used together with OpenID Connect. On the contrary, it can equally be used to build OAuth 2.0 federations or other protocols that depend on the dynamic exchange of Entity metadata.

An Entity Statement contains the information needed for the Entity that is the subject of the Entity Statement to participate in federation(s). An Entity Statement is always a signed JWT. The subject of the JWT is the Entity itself. The issuer of the JWT is the party that issued the Entity Statement. All Entities in a federation SHOULD be prepared to publish an Entity Statement about themselves (Entity Configuration). Entity Statement JWTs MUST be explicitly typed, by setting the typ header parameter to entity-statement+jwt. This is done to prevent cross-JWT confusion (see , section 3.11). An Entity Statement is composed of the following claims: REQUIRED. The Entity Identifier of the issuer of the statement. If the iss and the sub are identical, the issuer is making a statement about itself and this Entity Statement is an Entity Configuration. REQUIRED. The Entity Identifier of the subject. REQUIRED. The time the statement was issued. Its value is a JSON number representing the number of seconds from 1970-01-01T0:0:0Z as measured in UTC until the date/time. See for details regarding date/times in general and UTC in particular. REQUIRED. Expiration time on or after which the statement MUST NOT be accepted for processing. Its value is a JSON number representing the number of seconds from 1970-01-01T0:0:0Z as measured in UTC until the date/time. REQUIRED Conditional. A JSON Web Key Set (JWKS) representing the public part of the subject Entity's signing keys. The corresponding private key is used by Leaf Entities to sign Entity Statements about themselves, and intermediate entities to sign statements about other entities. The keys that can be found here are intended to verify the signature of the issued Entity Statements and Trust Marks and SHOULD NOT be used in other protocols. Keys to be used in other protocols, such as OpenID Connect, are conveyed in the metadata element of the respective Entity Statement. This claim is only OPTIONAL for the Entity Statement returned from an OP when the client is doing Explicit Registration. In all other cases it is REQUIRED. Every JWK in the JWK Set MUST have a Key ID (kid). It is RECOMMENDED that the Key ID be the JWK Thumbprint using the SHA-256 hash function of the key. OPTIONAL. The Entity Statement MAY be specifically created for an Entity. The Entity Identifier for that Entity MUST appear in this claim. OPTIONAL. Array of strings representing the Entity Identifiers of intermediate entities or Trust Anchors that MAY issue an Entity Statement about the issuer Entity. For all entities except for Trust Anchors that do not have any superiors this is REQUIRED and MUST NOT be the empty list []. This claim MUST be absent from an Entity Configuration issued by a Trust Anchor with no superiors. REQUIRED Conditional. JSON object including protocol specific metadata claims that represent the Entity's metadata. Each key of the JSON object represents a metadata type identifier, and each value MUST be a JSON object representing the metadata according to the metadata schema of that metadata type. An Entity Statement MAY contain multiple metadata statements, but only one for each metadata type. If the Entity Statement is an Entity Configuration, then the Entity Statement MUST contain a metadata claim. If iss and sub are not the same, then the Entity Statement MAY contain a metadata claim containing metadata asserted by a superior about the Entity identified by sub. If a metadata claim appears beside a metadata_policy claim in an Entity Statement then for each entity type, claims that appear in metadata MUST NOT appear in metadata_policy REQUIRED Conditional. JSON object that describes a metadata policy. A metadata_policy applies to the entity described in this Entity Statement as well as to all entities that appear beneath the entity described in this Entity Statement. This distinguishes metadata_policy from metadata which only applies to the subject itself. Each key of the JSON object represents a metadata type identifier, and each value MUST be a JSON object representing the metadata policy according to the metadata schema of that metadata type. An Entity Statement MAY contain multiple metadata policy statements, but only one for each metadata type. Only non Leaf Entities MAY contain a metadata_policy claim. Leaf entities MUST NOT contain a metadata_policy claim. OPTIONAL. JSON object that describes a set of Trust Chain constraints. There is more about this in . OPTIONAL. The crit (critical) Entity Statement claim indicates that extensions to Entity Statement claims defined by this specification are being used that MUST be understood and processed. It is used in the same way that crit is used for extension JWS header parameters that MUST be understood and processed. Its value is an array listing the Entity Statement claims present in the Entity Statement that use those extensions. If any of the listed extension Entity Statement claims are not understood and supported by the recipient, then the Entity Statement is invalid. Producers MUST NOT include Entity Statement claim names defined by this specification or names that do not occur as Entity Statement claim names in the Entity Statement in the crit list. Producers MUST NOT use the empty list [] as the crit value. OPTIONAL. The policy_language_crit (critical) Entity Statement claim indicates that extensions to the policy language defined by this specification are being used that MUST be understood and processed. It is used in the same way that crit is used for extension JSON Web Signature (JWS) header parameters that MUST be understood and processed. Its value is an array listing the policy language extensions present in the policy language statements that use those extensions. If any of the listed extension policy language extensions are not understood and supported by the recipient, then the Entity Statement is invalid. Producers MUST NOT include policy language names defined by this specification or names that do not occur in policy language statements in the Entity Statement in the policy_language_crit list. Producers MUST NOT use the empty list [] as the policy_language_crit value. OPTIONAL. A JSON array of objects, each representing a Trust Mark. Each JSON object MUST contain the following two claims and MAY contain other claims if needed. All the claims in the JSON object MUST have the same values as those contained in the Trust Mark JSON Web Token. The Trust Mark unique identifier. It MUST be the same value as the id claim contained in the Trust Mark JSON Web Token. A signed JSON Web Token that represents a Trust Mark. There is more about Trust Marks in . OPTIONAL. A Trust Anchor MAY use this claim to tell which combination of Trust Mark identifiers and issuers are trusted by the federation. This claim MUST be ignored if present in an Entity Configuration that describes an entity that is not a Trust Anchor. It is a JSON object with keys representing Trust Mark identifiers and values being an array of trusted entities representing the accreditation authority. If the value list bound to a Trust Mark identifier is empty it means that anyone can issue Trust Marks with that identifier. There is more about Trust Marks in . OPTIONAL. An OP MUST use this claim to tell the RP which Trust Anchor it chose to use when responding to an explicit client registration (Explicit Client Registration). The value of trust_anchor_id is the Entity Identifier of a Trust Anchor. The Entity Statement is signed using the private key of the issuer Entity, in the form of a JSON Web Signature (JWS). Implementations SHOULD support signature verification with the RSA SHA-256 algorithm because OpenID Connect Core requires support for it (alg value of RS256). Federations MAY also specify different mandatory-to-implement algorithms.

The following is a non-normative example of a trust_marks_issuers claim value:

The following is a non-normative example of an Entity Statement before serialization and adding a signature. The example contains a critical extension jti (JWT ID) to the Entity Statement and one critical extension to the policy language regexp (Regular expression).

In an OpenID Connect identity federation, entities that together build a Trust Chain can be categorized as: An Entity that represents a trusted third party. In an OpenID Connect identity federation, an RP or an OP. Neither a Leaf Entity nor a Trust Anchor. A Trust Chain begins with a Leaf Entity Configuration, has zero or more Entity Statements issued by intermediates about subordinates, including the Entity Statement issued by the Trust Anchor about the top-most intermediate (if there are intermediates) or the Leaf Entity (if there are no intermediates). The Trust Chain MAY end with the Entity Configuration of the Trust Anchor. If present, the Trust Anchor's Entity Configuration at the end of the chain contains configuration of the federation, such as the public keys of the Trust Anchor and other parameters considered within the trust evaluation. When this is this case, the Trust Chain consequentially contains the configuration of the federation at the date of the evaluation of the chain. A simple example: If we have an RP that belongs to organization A that is a member of federation F, the Trust Chain for such a setup will contain the following Entity Statements: an Entity Configuration about the RP published by the RP, an Entity Statement about the RP published by Organization A, an Entity Statement about Organization A published by federation F. A Trust Chain MUST always be possible to order such that: If we name the Entity Statements ES[0] (the Leaf Entity's Entity Configuration) to ES[i] (an Entity Statement issued by the Trust Anchor), i>0 then: The iss Entity in one Entity Statement is always the sub Entity in the next. ES[j]['iss'] == ES[j+1]['sub'], j=0,...,i-1 . There MUST always be a signing key carried in the jwks claim in ES[j] that can be used to verify the signature of ES[j-1], j=i,...,1 . It MUST be possible to verify the signature of ES[0] with one of the keys in ES[0]['jwks']. The signing key that MUST be used to verify ES[i] is distributed from the Trust Anchors to any Entity that needs to verify a Trust Chain in some secure out-of-band way not described in this document.

This specification does allow new metadata types to be defined, to support use cases outside OpenID Connect federations. The metadata type identifier will uniquely identify which metadata specification to utilize. The metadata document MUST be a JSON object. Beyond that there is no restriction. Metadata used in federations typically reuses existing metadata standards. If needed, the metadata schema is extended with additional properties relevant in a federated context.

For instance, for OpenID Connect federations, this specification uses metadata values from OpenID Connect Discovery 1.0 and OpenID Connect Dynamic Client Registration 1.0 and adds additional values used for federations. For OAuth2 federations, this specification uses metadata values from OAuth 2.0 Authorization Server Metadata as specified in . For both OpenID Connect and OAuth2 metadata the following additional properties are defined. OPTIONAL. A human-readable name representing the organization owning the Entity. If the owner is a physical person this MAY be, for example, the name of a person. Note that this information will be publicly available. OPTIONAL. A URI pointing to a signed JWT having the Entity's JWK Set as its payload. The JWT MUST be signed using a Federation Entity Key. A signed JWT can contain the following claims, all except keys defined in : REQUIRED. List of JWKs. REQUIRED. The "iss" (issuer) claim identifies the principal that issued the JWT. REQUIRED. This claim identifies the owner of the keys. It SHOULD be the same as the issuer. OPTIONAL. This claim identifies the time at which the JWT was issued. OPTIONAL. This claim identifies the time at which the JWT is no longer valid. There are more claims defined in ; of these, aud SHOULD NOT be used, since the issuer cannot know who the audience is. nbf and jti are deemed to not be very useful in this context and are therefore to be omitted.

The following is a non-normative example of a signed JWKS before serialization and adding a signature.

The following is a non-normative example of an RP's Entity Configuration:

OPTIONAL. JSON Web Key Set document, passed by value, as defined in . This parameter is intended to be used by participants that, for some reason, are unable to use the signed_jwks_uri parameter. One significant downside of jwks is that it does not enable key rotation (which signed_jwks_uri and jwks_uri do).

It is RECOMMENDED that an Entity Configuration use only one of jwks, jwks_uri, and signed_jwks_uri in its OpenID Connect or OAuth2 metadata. However, there may be circumstances in which is it desirable to use multiple JWK Set representations, such as when an Entity is in multiple federations and the federations have different policies about the JWK Set representation to be used. When multiple JWK Set representations are used, the keys present in each representation SHOULD be the same. Even if they are not completely the same at a given instant in time (which may be the case during key rollover operations), implementations MUST make them consistent in a timely manner.

The metadata type identifier is openid_relying_party. All parameters defined in Section 2 of OpenID Connect Dynamic Client Registration 1.0 are allowed in a metadata statement. In addition, the parameters defined in for OpenID Connect and OAuth2 entities and the following are added: REQUIRED. Array of strings specifying the client registration types the RP wants to use. Values defined by this specification are automatic and explicit.

The metadata type identifier is openid_provider. All parameters defined in Section 3 of OpenID Connect Discovery 1.0 are applicable. In addition, the parameters defined in for OpenID Connect and OAuth2 entities and the following are added: REQUIRED. Array specifying the federation types supported. Federation type values defined by this specification are automatic and explicit. OPTIONAL. URL of the OP's federation-specific Dynamic Client Registration Endpoint. If the OP supports explicit client registration as described in , then this claim is REQUIRED. OPTIONAL. A JSON Object defining the client authentications supported for each endpoint. The endpoint names are defined in the IANA "OAuth Authorization Server Metadata" registry . Other endpoints and authentication methods are possible, if made recognizable according to established standards and not in conflict with the operating principles of this specification. In OpenID Connect Core, no client authentication is performed at the authentication endpoint. Instead, because the request itself is authenticated. What it amounts to is that the OP maps information in the request (like the redirect_uri) to information it has gained on the client, through static or dynamic registration. If the map is successful, then the request is permitted to be processed. If the RP uses Automatic Registration, as defined in , the OP has no prior knowledge of the RP. Therefore, the OP must start by gathering information about the RP using the process outlined in . Once it has the RP's metadata, the OP can then verify the request in the same way as if it had known the RP's metadata beforehand. To make the request verification more secure we demand the use of a client authentication or verification method that proves that the RP is in possession of a key that appears in the RP's metadata. Client authentication methods are for instance: private_key_jwt. This authentication process is described in Section 9 of OpenID Connect Core 1.0. Note that if private_key_jwt is used, the audience of the signed JWT MUST be either the URL of the Authorization Server's Authorization Endpoint or the Authorization Server's Entity Identifier. Section 2.1 of . Section 2.2 of . A client verification method, on the other hand, is something like: This uses a Request Object as described in OpenID Connect Core 1.0. There is more about this in . Here the verification is based on the fact that the request object is signed with a key the RP is in control of. The claim value of this parameter is a JSON object with members representing authentication request methods and as values lists of client authentication/verification methods that can be used together with the authentication request method. Examples of authentication request methods are Authentication Request, as described in Section 3 of OpenID Connect Core 1.0, and Pushed Authorization Request (PAR), as described in . If AR is used, then a client verification method like request_object can be used. If PAR is used then client authentication methods like private_key_jwt, tls_client_auth and self_signed_tls_client_auth can be used. OPTIONAL. JSON array containing a list of the JWS signing algorithms (alg values) supported for the signature on the JWT used in the request_object contained in the request parameter of an authorization request or in the private_key_jwt of a pushed authorization request. This entry MUST be present if either of these authentication methods are specified in the request_authentication_methods_supported entry. No default algorithms are implied if this entry is omitted. Servers SHOULD support RS256. The value none MUST NOT be used. The following is a non-normative example of an OP's Entity Configuration:

The metadata type identifier is oauth_authorization_server. All parameters defined in Section 2 of RFC 8414 and are applicable. In addition, all the parameters defined for are applicable for this Entity.

The metadata type identifier is oauth_client. All parameters defined in Section 2 of RFC 7591 are applicable. In addition, the parameters defined in for OpenID Connect and OAuth2 entities are added.

The metadata type identifier is oauth_resource. In addition to those defined in draft-jones-oauth-resource-metadata , the parameters defined in for OpenID Connect and OAuth2 entities are added.

The metadata type identifier is federation_entity. All Entities in a federation MAY expose this metadata type. The Entities that expose federation API endpoints MUST expose this metadata type. The following properties are allowed: OPTIONAL. The fetch endpoint described in . Intermediate entities and Trust Anchors MUST publish a federation_fetch_endpoint. Leaf entities MUST NOT. OPTIONAL. The list endpoint described in . Intermediate entities and Trust Anchors MUST publish a federation_list_endpoint. Leaf entities MUST NOT. OPTIONAL. The resolve endpoint described in . Any federation Entity MAY publish a federation_resolve_endpoint. OPTIONAL. The Trust Mark status endpoint described in . Trust mark issuers SHOULD publish a federation_trust_mark_status_endpoint. OPTIONAL. A human-readable name representing the organization owning the Entity. If the owner is a physical person this MAY be, for example, the name of a person. Note that this information will be publicly available. It is RECOMMENDED that each federation Entity expose this claim. OPTIONAL. JSON array with one or more strings representing contact persons at the Entity. These MAY contain names, e-mail addresses, descriptions, phone numbers, etc. OPTIONAL. String. A URL that points to the logo of this Entity. The file containing the logo SHOULD be published in a format that can be viewed via the web. OPTIONAL. URL to documentation of conditions and policies relevant to this Entity. OPTIONAL. URL to a generic home page representing this Entity. Example

An Entity can publish metadata policies pertaining to subordinate entities of a specific type. Entity type identifiers specified in this document can be found in .

A metadata_policy for a specific entity type has the following structure: It consists of one or more policy entries. Each policy entry applies to one metadata parameter, such as id_token_signed_response_alg. Each policy entry consists of one or more operators, which can be value modifiers or value checks. An operator can only appear once in a policy entry. It SHOULD be noted that claim names without language tags are different from the same claim but with language tags.

An example of a policy entry:

Which fits into a metadata policy like this:

Value modifiers are: Disregarding what value the parameter had, if any, the parameter's value will be set to the operator's value. Adds the value or values specified to the metadata parameter. If any of the specified values are already present as values of the parameter, they will not be added. If the parameter has no value, then the parameter is initialized with the specified value(s). If no value is assigned to this parameter, then the parameter's value will be set to the operator's value(s). Value checks are: The value of the parameter MUST be one of the ones listed in this directive. The resulting value of the parameter will be the intersection of the values in the directive and the values of the parameter. The values of the parameter MUST contain the ones in the directive. We define superset in a mathematical way, that is, equality is included. If true, then the parameter MUST have a value. Note that when a parameter is defined to be a space-separated list of strings like for scope in subset_of, superset_of and default are still expressed as lists of strings. This language from also applies for such space-separated lists of strings: "If the value contains multiple space-delimited strings, their order does not matter, and each string adds an additional access range to the requested scope."

As stated, a policy entry can contain one or more operators. Not all operators are allowed to appear together in a policy entry. subset_of and superset_of applies to parameters that can have more than one value (for instance, contacts) while one_of applies to parameters that can only have one value (for instance, id_token_signed_response_alg). This means that one_of cannot appear beside subset_of/ superset_of in a policy entry. value overrides everything else. So having value together with any other operator (except for essential) does not make sense. Other restrictions are: If subset_of and superset_of both appear as operators, then the list of values in subset_of MUST be a superset of the values in superset_of. If add appears in a policy entry together with subset_of then the value/values of add MUST be a subset of subset_of. If add appears in a policy entry together with superset_of then the values of add MUST be a superset of superset_of. If default appears in a policy entry together with subset_of then the values of default MUST be a subset of subset_of. If default appears in a policy entry together with superset_of then the values of default MUST be a superset of superset_of. If add appears in a policy entry together with one_of then the value of add MUST be a member of one_of. If default appears in a policy entry together with one_of then the value default MUST be a member of one_of.

If there is more than one metadata policy in a Trust Chain, then the policies MUST be combined before they are applied to the metadata statement. Using the notation we have defined in , policies are combined starting with ES[i] and then adding the policies from ES[j] j=i-1,..,1 before applying the combined policy to the Entity's metadata. After each combination, the policy for each parameter MUST adhere to the rules defined in .

The result of merging the values of two subset_of operators is the intersection of the operator values. The result of merging the values of two one_of operators is the intersection of the operator values. The result of merging the values of two superset_of operators is the union of the operator values. The result of merging the values of two add operators is the union of the values. Merging two value operators is NOT allowed unless the two operator values are equal. Merging two default operators is NOT allowed unless the two operator values are equal. If a superior has specified essential=true, then a subordinate cannot change that. If a superior has specified essential=false, then a subordinate is allowed to change that to essential=true. If a superior has not specified essential, then a subordinate can set essential to true or false.

Once combining the Metadata policies has been accomplished, the next step is to apply the combined policy to the metadata. Doing that, one follows these steps for each parameter in the policy. If there is a value operator in the policy, its value would be applied without any additional action. Add whatever value is specified in an add operator. If the parameter still has no value, apply the default if there is one. Do the essential check. If essential is missing as an operator, essential is to be treated as if set to false. If essential is defined to be true, then the claim MUST have a value by now. Otherwise applying the operator MUST fail. Do the other checks. Verify that the value is one_of or that the values are subset_of/superset_of. If the parameter values do not fall within the allowed boundaries, applying the operator MUST fail. If more than one value is specified in one_of and the Leaf Entity does not specify the metadata claim for which one_of is referred to, it is an implementation choice which of the values in the one_of array will be used.

A federation's policy for OAuth2 clients:

An organization's policy for OAuth2 clients:

The combined metadata policy then becomes:

If applying policies to a metadata statement results in incorrect metadata, then such a metadata statement MUST be regarded as broken and MUST NOT be used.

There might be parties that want to extend the policy language defined here. If that happens then the rule is that if software compliant with this specification encounters a keyword it does not understand, it MUST ignore it unless it is listed in a policy_language_crit list, as is done for JWS header parameters with the crit parameter. If the policy language extension keyword is listed in the policy_language_crit list and not understood, then the metadata MUST be rejected.

The following is a non-normative example of a set of policies being applied to an RP's metadata.

The RP's metadata:

The federation's policy for RPs:

The organization's policy for RPs:

The metadata for the Entity in question, after applying the policies above, would then become:

A constraint specification can contain the following claims: OPTIONAL. Integer. The maximum number of Entity Statements between this Entity Statement and the last Entity Statement in the Trust Chain. OPTIONAL. JSON object. Restriction on the Entity Identifiers of the entities below this Entity. The behavior of this claim mimics what is defined in Section 4.2.1.10 of . Restrictions are defined in terms of permitted or excluded name subtrees. OPTIONAL. Array of strings. Each string is an entity type. Entity type identifiers specified in this document can be found in . This constraint restricts what types of Leaf Entities that MAY appear beneath the entity described in this Entity Statement. The following is a non-normative example of such a specification:

If a subordinate Entity Statement contains a constraint specification that is more restrictive than the one in effect, then the more restrictive constraint is in effect from here on. If a subordinate Entity Statement contains a constraint specification that is less restrictive than the one in effect, then it MUST be ignored.

The max_path_length constraint specifies the maximum number of Entity Statements a Trust Chain can have between the Entity Statement that contains the constraint specification and the Leaf's Entity Statement. A max_path_length constraint of zero indicates that no intermediaries MAY appear between this Entity and the Leaf Entity. Where it appears, the max_path_length constraint MUST have a value that is greater than or equal to zero. Not adding a max_path_length constraint just means that there is no additional constraints apart from those already in effect. Assuming that we have a Trust Chain with four Entity Statements: Entity Configuration of the Leaf Entity (LE) Entity Statement by Intermediate 1 (I1) about LE Entity Statement by Intermediate 2 (I2) about I1 Entity Statement by Trust Anchor (TA) about I2 Then the Trust Chain fulfills the constraints if for instance: The TA specifies a max_path_length that is equal to or bigger than 2. TA specifies max_path_length of 2, I2 specifies max_path_length of 1, and I1 sets no max_path_length constraint. Neither TA nor I2 specifies any max_path_length constraint while I1 sets max_path_length to 0. The Trust Chain does not fulfill the constraints if for instance the: TA's Entity Statement has set the max_path_length to 1.

The naming_constraints member specifies a namespace within which all subject Entity Identifiers in subordinate Entity Statements in a Trust Chain MUST be located. Restrictions are defined in terms of permitted or excluded name subtrees. Any name matching a restriction in the excluded claim is invalid regardless of information appearing in the permitted claim. Domain name constraints are as specified in Section 4.2.1.10 of . As stated there, a domain name constraint MUST be specified as a fully qualified domain name and MAY specify a host or a domain. Examples are "host.example.com" and ".example.com". When the domain name constraint begins with a period, it MAY be expanded with one or more labels. That is, the domain name constraint ".example.com" is satisfied by both host.example.com and my.host.example.com. However, the domain name constraint ".example.com" is not satisfied by "example.com". When the domain name constraint does not begin with a period, it specifies a host.

The allowed_leaf_entity_types constraint specifies what entity types are possible for Leaf Entities beneath an entity. If there is no allowed_leaf_entity_types defined it means that any type is allowed.

Technically, Trust Marks as used by this specification are signed JWTs. The Trust Marks are signed by a federation accredited authority. The validation of such a signed statement is performed in the same way that an Entity Configuration is validated. To make this validation possible the key used by the Trust Mark Issuer to sign Trust Marks MUST be one of its the Federation Entity Keys. Note that a federation MAY allow an Entity to self-sign some Trust Marks. Trust Mark JWTs MUST be explicitly typed, by setting the typ header parameter to trust-mark+jwt. This is done to prevent cross-JWT confusion (see , section 3.11).

These are the Trust Mark Claims: REQUIRED. String. The issuer of the Trust Mark. REQUIRED. String. The Entity this Trust Mark applies to. REQUIRED. String. An identifier of the Trust Mark. The Trust Mark id MUST be collision-resistant across multiple federations. It is RECOMMENDED that the id value is built using the unique URL that uniquely identifies the federation, or the trust framework, within which it was issued. This is required to prevent Trust Marks issued in different federations from having colliding identifiers. REQUIRED. Number. When this Trust Mark was issued. Expressed as Seconds Since the Epoch, per . OPTIONAL. String. A URL that points to a mark/logo that the subject is allowed to display to a user of the Entity. OPTIONAL. Number. When this Trust Mark is not valid anymore. Expressed as Seconds Since the Epoch, per . If not present, it means that the Trust Mark is valid forever. OPTIONAL. String. URL that points to information connected to the issuance of this Trust Mark. Other claims MAY be used in conjunction with the claims outlined above. The claim naming recommendations outlined in Section 5.1.2 of OpenID Connect Core 1.0 apply.

An Entity SHOULD NOT try to validate a Trust Mark until it knows which Trust Anchors it wants to use. Validating a Trust Mark issuer follows the procedure set out in . Validating the Trust Mark itself is described in Note that the Entity representing the accreditation authority SHOULD be well known and trusted for a given Trust Mark identifier. A Trust Anchor MAY publish a list of accreditation authorities of Trust Marks that SHOULD be trusted by other Entities. A Trust Anchor uses the trust_marks_issuers claim in its Entity Configuration to publish this information. For other externally issued Trust Marks, it is an out-of-band process to define and announce accreditation authorities to other entities and it is left to the discretion of the receiving party to assign an appropriate level of trust to such Trust Marks.

A non-normative example of a Trust Mark claim inside an Entity Configuration is:

An example of a decoded Trust Mark issued to a RP, attesting the conformance to a national public service profile:

An example of a decoded Trust Mark issued to a RP, attesting its conformance to the rules for data management of underage users:

An example of a Trust Mark attesting a stipulation of an agreement between a RP and an Attribute Authority:

An example of a Trust Mark asserting conformance to a security profile:

An example of a decoded self-signed Trust Mark:

An example of a third-party accreditation authority:

The Entity Configuration of every participant SHOULD be exposed at a well-known endpoint. The configuration endpoint is found using the Well-Known URIs specification, with the suffix openid-federation. The scheme, host, and port are taken directly from the Entity Identifier combined with the following path: /.well-known/openid-federation. If the Entity Identifier contains a path, it is concatenated after /.well-known/openid-federation in the same manner that path components are concatenated to the well-known identifier in the OAuth 2.0 Authorization Server Metadata specification. Of course, in real multi-tenant deployments, in which the Entity identifier might be of the form https://multi-tenant-service.example.com/my-tenant-identifier the tenant is very likely to not have control over the path https://multi-tenant-service.example.com/.well-known/openid-federation/my-tenant-identifier whereas it is very likely to have control over the path https://multi-tenant-service.example.com/my-tenant-identifier/.well-known/openid-federation. Therefore, if using the configuration endpoint at the URL with the tenant path after the well-known part fails, it is RECOMMENDED that callers retry at the URL with the tenant path before the well-known part (even though this violates ). Entities SHOULD make an Entity Configuration document available at the configuration endpoint. There is only one exception to this rule and that is for an RP that only does Explicit Registration. Since it posts the Entity Configuration to the OP during client registration, the OP has everything it needs from the RP.

An Entity Configuration document MUST be queried using an HTTP GET request at the previously specified path. The requesting party would make the following request to the Entity https://openid.sunet.se to obtain its configuration information:

The response is an Entity Configuration, as described in . If the Entity is an intermediate Entity or a Trust Anchor, the response MUST contain metadata for a federation Entity (federation_entity). A successful response MUST use the HTTP status code 200 and the content type set to application/entity-statement+jwt, to make it clear that the response contains a signed Entity Statement. In case of an error, the response SHOULD be produced in accordance with what is defined in . The following is a non-normative example response from an intermediate Entity, before serialization and adding a signature:

The federation endpoints of an Entity can be found in the configuration response as described in or by other means.

All Entities that are expected to publish Entity Statements about other entities MUST expose a Fetch endpoint. Fetching Entity Statements is performed to collect Entity Statements one by one to gather Trust Chains. To fetch an Entity Statement, an Entity needs to know the identifier of the Entity to ask (the issuer), the fetch endpoint of that Entity and the identifier of the Entity, that is the subject of the statement.

The request MUST be an HTTP request using the GET method and the https scheme to a resolved federation endpoint with the following query string parameters, encoded in application/x-www-form-urlencoded format. OPTIONAL. The Entity Identifier of the issuer from which the Entity Statement issued. Because of the normalization of the URL, multiple issuers MAY resolve to a shared fetch endpoint. This parameter makes it explicit exactly which issuer the Entity Statements must come from. Without this parameter, the issuer matches to the Entity at which the request was made. OPTIONAL. The Entity Identifier of the subject for which the Entity Statement is issued. If this parameter is left out, it is considered to be the same as the issuer and would indicate a request for a self-signed statement.

The following is a non-normative example of an API request for an Entity Statement:

A successful response MUST use the HTTP status code 200 and the content type set to application/entity-statement+jwt, to make it clear that the response contains a signed Entity Statement. If it is a negative response, it will be a JSON object and the content type MUST be set to application/json. See more about error responses in .

The following is a non-normative example of a response, before serialization and adding a signature:

An Entity MAY use the resolve endpoint to fetch resolved metadata and Trust Marks for an Entity as seen/trusted by the resolver. The resolver is supposed to fetch the subject's Entity Configuration, collect a Trust Chain that starts with the subject's Entity Statement and ends with the specified Trust Anchor, verify the Trust Chain and then apply all the policies present in the Trust Chain to the Entity Statements metadata. The resolver is also expected to verify that the present Trust Marks are active. If it finds Trust Marks that are not active, then those should be left out of the response set.

The request MUST be an HTTP request using the GET method and the https scheme to a resolve endpoint with the following query string parameters, encoded in application/x-www-form-urlencoded format. REQUIRED. The Entity Identifier of the Entity whose resolved data is requested. REQUIRED. The Trust Anchor that the remote peer MUST use when resolving the metadata. The value is an Entity identifier. OPTIONAL. A specific metadata type to resolve. In this document, we use the metadata types listed in . If no value is given, then all metadata types are expected to be returned.

The following is a non-normative example of a resolve request:

A successful response MUST use the HTTP status code 200 and the content type set to application/resolve-response+jwt, containing resolved metadata and verified Trust Marks. The resolve response JWT MAY also contain the sequence of the Entity Statements that compose the Trust Chain, sorted as shown in . The key used by the resolver to sign the response MUST be one of its Federation Entity Keys. The response SHOULD contain the aud claim only if the requesting client is authenticated. These are the resolve response Claims: REQUIRED. String. URL corresponding to the federation's Entity Identifier of the issuer of the resolve response. REQUIRED. String. URL corresponding to the federation's Entity Identifier of the subject for which the resolve response refers to. REQUIRED. Number. When this resolution was issued. Expressed as Seconds Since the Epoch, as defined in . REQUIRED. Number. When this resolution is not valid anymore. Expressed as Seconds Since the Epoch, as defined in . It MUST correspond to the exp value of the Trust Chain from which the resolve response was derived. REQUIRED. JSON object containing the resolved subject metadata, according to the requested type and expressed in the metadata format defined in . OPTIONAL. A JSON array of objects, each representing a Trust Mark, as defined in . OPTIONAL. Array containing the sequence of the statements that compose the Trust Chain starting with the subject and ending with the selected Trust Anchor, sorted as shown in . Other claims MAY be used in conjunction with the claims outlined above. The claim naming recommendations outlined in Section 5.1.2 of OpenID Connect Core 1.0 apply. If the response is negative, the response SHOULD be produced in accordance with what is defined in .

The following is a non-normative example of a response, before serialization and adding a signature:

The basic assumption of this specification is that an entity should have direct trust in no one except the Trust Anchor and its own capabilities. However, Entities may establish a kind of transistive trust in other entities. For example, the Trust Anchor states who its subordinates are and Entities may choose to trust these. If a party uses the resolve service of another Entity to obtain federation data, it is trusting the resolver to perform the validation of the cryptographically protected metadata correctly and to provide it with authentic results.

An Entity MAY query another Entity for a list of all the entities immediately subordinate to that Entity and about which that Entity is prepared to issue statements about. (In some cases, this MAY be a very large list.)

The request MUST be an HTTP request using the GET method and the https scheme to a list endpoint with the following query parameters, encoded in application/x-www-form-urlencoded format. OPTIONAL. If the responder knows the metadata type identifier of all its subordinates, the result MUST be filtered accordingly. If the responder does not support this feature it MUST use the HTTP status code 400 and set the content type to application/json, with the error code set to unsupported_parameter.

The following is a non-normative example of an API request for a list of entities:

A successful response MUST use the HTTP status code 200 and the content type set to application/json, containing a JSON list with the known Entity Identifiers. If the response is negative, the response should be produced in accordance with what is defined in .

The following is a non-normative example of a response:

This is to allow an Entity to check whether a Trust Mark is still active or not. The query MUST be sent to the Trust Mark issuer.

The request MUST be an HTTP request using the POST method and the https scheme to a status endpoint with the following parameters, encoded in application/x-www-form-urlencoded format. OPTIONAL. The ID of the Entity to which the Trust Mark was issued. OPTIONAL. Identifier of the Trust Mark. OPTIONAL. When the Trust Mark was issued. If iat is not specified and the trust issuer have issued several Trust Marks with the id specified in the request to the Entity identified by sub; the last one is assumed. OPTIONAL. The whole Trust Mark. If trust_mark is used then sub and id are not needed. If trust_mark is not used, then sub and id are required.

The following is a non-normative example of a request using sub and id:

A successful response MUST use the HTTP status code 200 and the content type set to application/json, to make it clear that the response contains a JSON object containing the claim below. If the response is negative, the response SHOULD be produced in accordance with what is defined in . REQUIRED. Boolean. Whether the Trust Mark is active or not.

The following is a non-normative example of a response:

The Trust Anchor MAY publish its expired federation public keys at the endpoint /.well-known/openid-federation-historical-jwks . The purpose of this endpoint is to provide the list of keys previously used by the Trust Anchor in order to provide non-repudiation of statements signed by the Trust Anchor after key rotation. The historical keys of the Trust Anchor guarantees participants the verifiability of the Trust Chains even when the federation keys are changed.

The request MUST be an HTTP request using the GET method and the https scheme to the well known endpoint openid-federation-historical-jwks.

The following is a non-normative example of a request:

A successful response MUST use the HTTP status code 200 and the content type set to application/jwk-set+jwt. The response is a signed JWT containing the following paramenters: REQUIRED. String. URL corresponding to the Trust Anchor Entity Identifier. REQUIRED. Number. When the signed JWT was issued, using the time format defined for the iat claim in . REQUIRED. A JSON Web Key Set (JWKS) representing the public part of the Trust Anchor signing keys. The JWK Set in the jwks claim, in addition to what is defined in JSON Web Key Set (JWKS) also adopts the following claims: REQUIRED. Parameter is used to match a specific key. It is RECOMMENDED that the Key ID be the JWK Thumbprint using the SHA-256 hash function of the key. REQUIRED. Time at which the JWK was issued, using the time format defined for the iat claim in . REQUIRED. Expiration time, using the time format defined for the exp claim in . on or after which the JWK MUST NOT be considered as valid.

The following is a non-normative example of a response, before serialization and adding a signature:

The Federation Historical Keys endpoint solves the problem of verifying historical Trust Chains when the Trust Anchor Federation Entity Keys are changed. The Federation Historical Keys endpoint publishes the list of public keys used in the past by Trust Anchor. The given public keys are needed by all the Entities to verify the Trust Chains evaluated in the past with the Federation Entity Keys not published anymore in the Trust Anchor's Entity Configuration. Federation Historical Keys endpoint response contains a signed JWT that attests all the expired Trust Anchor Keys. On the basis of the attributes contained in the Entity Statements that form a Trust Chain, it MAY be also possible to verify the non-federation public keys used in the past by a Leaf for the signature operations related to the OpenID Connect requests and responses. For example an Entity Statement issued for a Leaf MAY also include the jwks claim, respectively in the claims metadata or metadata_policy. A simple example: In the following Trust Chain the Federation Intermediary attests the Leaf's OpenID Connect jwks in the Entity Statement issued for the Leaf. The result is a Trust Chain that contains the Leaf's OIDC Core jwks, needed to verify historical signature on request objects, ID Tokens and any other signed JWT issued by the Leaf, if it is an RP or an OP. an Entity Configuration about the RP published by the RP, an Entity Statement about the RP published by Organization A, with the claim jwks contained in metadata or metadata_policy attesting the Leaf's OpenID Core jwks. an Entity Statement about Organization A published by Federation F.

If the request was malformed, or some error occurred during processing of the request, the response body SHOULD be a JSON object with the content type set to application/json. In compliance with , the following standardized error format SHOULD be used: REQUIRED. One of the following error codes SHOULD be used. The request is incomplete or does not comply with current specifications. The HTTP response status code SHOULD be 400 (Bad Request). The Client can not be authorized or is not a valid participant of the federation. The HTTP response status code SHOULD be 401 (Unauthorized). The requested Entity Identifier is "not found". The HTTP response status code SHOULD be 404 (Not Found). The server encountered an unexpected condition that prevented it from fulfilling the request. The HTTP response status code SHOULD be one in the 5xx range, like 500 (Internal Server Error). The server hosting the federation endpoint is currently unable to handle the request due to a temporary overloading or maintenance. The HTTP response status code SHOULD be 503 (Service Unavailable). The server does not support the requested parameter. REQUIRED. A human-readable short text describing the error.

The following is a non-normative example of an error response:

An Entity (e.g., the consumer) that wants to establish trust with a remote peer MUST have the remote peer's Entity Identifier and a list of Entity Identifiers of Trust Anchors together with the public version of their signing keys. The consumer will first have to fetch sufficient Entity Statements to establish at least one chain of trust from the remote peer to one or more of the configured Trust Anchors. After that the Entity MUST validate the Trust Chains independently, and -- if there are multiple valid Trust Chains and if the application demands it -- choose one.

Depending on the circumstances, the consumer MAY either be handed the remote peer's Entity Configuration, or it may have to fetch it by itself. If it needs to fetch it, it will use the process described in based on the Entity Identifier of the remote peer. The next step is to iterate through the list of intermediates listed in authority_hints, ignoring the authority hints that end in an unknown Trust Anchor, requesting an Entity Configuration from each of the intermediates. If the received Entity Configuration contains an authority hint this process is repeated. With the list of all intermediates and the Trust Anchor, the respective federation API (see ) is used to fetch Entity Statements about the intermediates and the remote peer. Note: The consumer SHOULD NOT attempt to fetch Entity Statements it already has fetched during this process (loop prevention). A successful operation will return one or more lists of Entity Statements. Each of the lists terminating in a self-signed Entity Statement is issued by a Trust Anchor. If there is no path from the remote peer to at least one of the trusted Trust Anchors, then the list will be empty and there is no way of establishing trust in the remote peer's information. How the consumer deals with this is out of scope for this specification.

As described in , a Trust Chain consists of an ordered list of Entity Statements. So whichever way the consumer has acquired the set of Entity Statements, it MUST now verify that it is a proper Trust Chain using the rules laid out in that section. To validate the chain, the following MUST be done: For each Entity Statement ES[j] j=i,..,0: Verify that the statement contains all the required claims. Verify that iat has a value in the past. Verify that exp has a value that is in the future. For j=0 verify that iss == sub. For j=0,...,i-1: Verify that ES[j]['iss'] == ES[j+1]['sub'] For j=0,...,i-1: Verify the signature of ES[j] using a public key carried in ES[j+1]['jwks']. For j == 0 verify the signature of ES[0] using a public key carried in ES[0]['jwks']. For j == i: verify that a) the issuer matches the configured identifier of a Trust Anchor and b) its signature is valid with the likewise configured public key of said Trust Anchor. Verifying the signature is a much more expensive operation than verifying the correctness of the statement and the timestamps. An implementer MAY therefore choose to not verify the signature until all the other checks have been done. Consumers MAY cache Entity Statements or signature verification results for a given time until they expire, per . Note that the second bullet point means that, at each step in the Trust Chain resolution, it MUST be verified that the signing JWK is also present in the jwks statement claim issued by the superior.

If multiple valid Trust Chains are found, the consumer will need to decide on which one to use. One simple rule would be to prefer a shorter chain over a longer one. Consumers MAY follow other rules according to local policy.

Each Entity Statement in a Trust Chain is signed and MUST have an expiration time (exp) set. The expiration time of the whole Trust Chain is set to the minimum value of (exp) within the chain.

This specification allows for a smooth process of updating metadata and public keys. As described above in , each Trust Chain has an expiration time. A consumer of metadata using this specification MUST support refreshing a Trust Chain when it expires. How often a consumer SHOULD re-evaluate the Trust Chain depends on how quickly the consumer wants to find out that something has changed in the Trust Chain.

If a Leaf Entity publishes its public keys in the metadata part using jwks, setting an expiration time on the self-signed Entity Statement can be used to control how often the receiving Entity is fetching an updated version of the public key.

A Trust Anchor MUST publish an Entity Configuration about itself. The Trust Anchor SHOULD set a reasonable expiration time on that Statement, such that the consumers will re-fetch the Entity Configuration at reasonable intervals. If the Trust Anchor wants to roll over its signing keys it would have to: Add the new keys to the jwks representing the Trust Anchor's signing keys. Keep signing the Entity Configuration and the Entity Statements using the old keys for a long enough time period to allow all subordinates to have gotten access to the new keys. Switch to signing with the new keys. After a reasonable time period remove the old keys. What is regarded as a reasonable time is dependent on the security profile and risk assessment of the Trust Anchor.

It is RECOMMENDED that Federation Operators provide a means of retrieving the public keys for Trust Anchors that it administers that is independent of the Entity Configurations for those Trust Anchors. This is intended to provide redundancy in the eventuality of the compromise of the Web PKI infrastructure underlying retrieval of public keys from Entity Configurations. The keys retreived via the independent mechanism specified by the Federation Operator SHOULD be be compared to those retreived via the Trust Anchor's Entity Configuration. If they do not match, both SHOULD be retrieved again. If they still do not match, it is indicative of a security or configuration problem. The appropriate remediation steps in that eventuality SHOULD be specified by the Federation Operator.

Since the consumers are expected to check the Trust Chain at regular, reasonably frequent times, this specification does not specify a standard revocation process. Specific federations MAY make a different choice and will then have to add such a process.

This section describes how the trust framework in this specification is used to establish trust between an RP and an OP that have no explicit configuration or registration in advance. There are two alternative approaches to establish trust between an RP and an OP, which we call Automatic and Explicit Registration. Members of a federation or a community SHOULD agree upon which one to use. While implementations should support both methods, deployments MAY choose to disable the use of one of them. Independent of whether the RP uses Automatic or Explicit Registration, the way that the RP learns about the OP is the same. It will use the procedure that is described in .

Automatic Registration enables an RP to make Authentication Requests without a prior registration step with the OP. The OP resolves the RP's Entity Configuration from the Client ID in the Authentication Request, following the process defined in . Automatic Registration has the following characteristics: In all interactions with the OP, the RP employs its Entity Identifier as the Client ID. The OP retrieves the RP's Entity Configuration from a URL derived from the Entity Identifier, as described in . Since there is no registration step prior to the Authentication Request, the RP cannot be supplied with a Client Secret. Instead, the Automatic Registration model requires the RP to use asymmetric cryptography to authenticate its requests. The OP MUST publish that it supports a request authentication method using the metadata claim request_authentication_methods_supported.

The Authentication Request is performed by passing a Request Object by value as described in Section 6.1 in OpenID Connect Core 1.0 or using pushed authorization as described in Pushed Authorization Requests.

In the case where a Request Object is used, the value of the request parameter is a JWT whose Claims are the request parameters specified in Section 3.1.2 in OpenID Connect Core 1.0. The JWT MUST be signed and MAY be encrypted. The following restrictions apply to the JWT: REQUIRED. The Audience (aud) value MUST be or include the OP's Issuer Identifier URL. REQUIRED. The claim iss MUST contain the client identifier. MUST NOT be present. This together with the value of aud SHOULD make reuse of the statement for private_key_jwt client authentication not feasible. REQUIRED. JWT ID. A unique identifier for the JWT, which can be used to prevent reuse of the token. These tokens MUST only be used once, unless conditions for reuse were negotiated between the parties; any such negotiation is beyond the scope of this specification. REQUIRED. Expiration time on or after which the JWT MUST NOT be accepted for processing. OPTIONAL. Time at which the JWT was issued. OPTIONAL. Array containing the sequence of the statements that compose the Trust Chain between the RP that makes the request and the selected Trust Anchor, sorted as shown in . Due to the large size of a Trust Chain it could be necessary to use the HTTP POST method, request_uri or Pushed Authorization Request for the request.

The following is a non-normative example of the Claims in a Request Object before base64url encoding and signing:

The following is a non-normative example of an Authentication Request using the request parameter (with line wraps within values for display purposes only):

When the OP receives an incoming Authentication Request, the OP supports OpenID Connect Federation, the incoming Client ID is a valid URL, and the OP does not have the Client ID registered as a known client, then the OP SHOULD resolve the Trust Chains related to the requestor. An RP MAY present to the OP a Trust Chain related to itself, using the trust_chain parameter. If the OP doesn't have a valid registration for the RP or its registration has expired, the OP MAY use the received Trust Chain as a hint which path to take from the Leaf Entity to the Trust Anchor. The OP MAY evaluate the statements in the trust_chain to make its Federation Entity Discovery procedure more efficient, especially if the RP shows more than a single authority hint in its Entity Configuration. If the OP already has a valid registration for the RP it MAY use the received Trust Chain to update the RP's registration. Whenever the OP uses a Trust Chain submitted by an RP, the OP MUST fully verify it, with every statement contained in it. A Trust Chain may be relied upon by the OP because it has validated all of its statements. This is true whether these statements are retrieved from their URLs or whether they are provided via the trust_chain parameter. If the RP doesn't include the trust_chain in the request or the OP doesn't support this feature, it then MUST validate the possible Trust Chains, starting with the RP's Entity Configuration as described in , and resolve the RP metadata with type openid_relying_party. The OP SHOULD furthermore consider the resolved metadata of the RP, and verify that it complies with the client metadata specification in OpenID Connect Dynamic Client Registration 1.0. Once the OP has the RP's metadata, it can verify that the client was actually the one sending the Authentication Request by verifying the signature of the Request Object using the key material the client published through its metadata (underneath metadata/openid_relying_party).

Pushed Authorization Requests provides an interoperable way to push the payload of a Request Object directly to the AS in exchange for a request_uri. When it comes to request authentication, the applicable methods are four: Using a JWT for Client authentication as described for private_key_jwt in Section 9 of OpenID Connect Core 1.0. mTLS as described in Section 2.2 of based on self-signed certificates. In this case the self-signed certificate MUST be present as the value of an x5c claim for one key in the JWK Set describing the RP's keys. mTLS as described in Section 2.1 of based on public key infrastructure (PKI). A request object as described in Section 6 of OpenID Connect Core 1.0 Note that if mTLS is used, TLS client authentication MUST be configured and, in case of self-signed certificates, the server must omit the certificate chain validation. Using the example above, a request could look like this:

All the assumptions and requirements already defined in also apply to Pushed Authorization Requests. Once the OP has the RP's metadata, it can verify the client using the keys published underneath the metadata/openid_relying_party element. This is where it diverges depending on which client authentication method was used. If this method is used, then the OP will try to verify the signature of the signed JWT using the key material published by the RP in its metadata. If the authentication is successful, then the registration is regarded as correct. If mTLS is used and the certificate used was not self-signed, then the Subject Alternative Name of the certificate MUST match the Entity Identifier of the RP. If mTLS is used and the certificate used is a self-signed certificate, then the certificate MUST be present as the value of an x5c claim for one key in the JWK Set describing the RP's keys.

If the OP fails to establish trust with the RP, it SHOULD use an appropriate error code, and an error_description that aids the RP to understand what is wrong. In addition to the error codes defined in Section 3.1.2.6 of OpenID Connect Core, this specification also defines the following error codes: No trusted Trust Anchor could be found. Trust chain validation failed.

The following is a non-normative example error response:

Note that we anticipate that Automatic Registration could be employed for use cases beyond OpenID Connect Federation. If used with pure OAuth 2.0, for instance, Automatic Registration would occur during Authorization Requests. If for a particular such use case the client does not need to be securely identified, then the requirement for signed requests could be relaxed.

This method involves performing an explicit registration step for a new client before the RP interacts with an OP for the first time, similarly to the process specified by OpenID Connect Dynamic Client Registration 1.0, but where the client registration request contains the Entity Configuration or an entire Trust Chain.

An OP that supports OpenID Dynamic Client Registration as extended by this specification, signals this by having the claim federation_registration_endpoint in the OP's metadata. Given that the OP supports Explicit Registration, the RP progresses as follows: Once it has a list of acceptable Trust Anchors to the OP, it MUST choose the subset it wants to progress with. This subset can be as small as one Trust Anchor, but it can also contain more than one. Using this subset of Trust Anchors, the RP will choose a set of authority_hints from the hints that are available to it. Each hint MUST, when used as a starting point for Trust Chain collection, lead to at least one of the Trust Anchors in the subset. The RP will now construct its Entity Configuration where the metadata statement chosen is influenced by the OP's metadata and where the authority_hints included are picked by the process described above. The RP MAY include its Entity Configuration in a Trust Chain regarding itself. The Registration Request will contain an array containing the sequence of the statements that compose the Trust Chain between the RP that makes the request and the selected Trust Anchor. The Entity Configuration, or the entire Trust Chain, is sent, using POST, to the federation_registration_endpoint defined in this document. The content type of the Registration Request MUST be set to application/entity-statement+jwt if it contains only the Entity Configuration of the requestor. Otherwise if it contains the Trust Chain, the content type of the Registration Request MUST be set to application/trust-chain+json.

After the OP receives the registration request, it checks if this contains an Entity Configuration or an entire Trust Chain. If the request contains an Entity Configuration, the OP collects and evaluates the Trust Chains starting with the authority_hints in the Entity Configuration of the requestor. After it has verified at least one Trust Chain it MUST verify that the signature on the received registration request is correct. If the OP finds more than one acceptable Trust Chain, it MUST choose one Trust Anchor from those chains as the one it will proceed with. If the request contains a Trust Chain, the OP MAY evaluate the statements in the Trust Chain to make its Federation Entity Discovery procedure more efficient, especially if the RP shows more than a single authority hint in its Entity Configuration. At this point, if there already exists a client registration under the same Entity Identifier then that registration MUST be regarded as invalid. Note that key material from the previous registration SHOULD be kept to enable verifying signatures or decrypting archived data. The OP will now construct metadata policies and assertions that, if applied to the RP's metadata statement, will result in metadata that the OP finds acceptable. Note that the Client ID the OP chooses does not have to be the same as the Entity Identifier of the RP. To the Entity Statement it MUST add a trust_anchor_id claim, containing the Trust Anchor chosen in step 2. It will sign and return the registration response (a signed Entity Statement) to the RP. A successful response MUST use the HTTP status code 200 and the content type set to application/jose. If the response is negative, the response SHOULD be produced in accordance with what is defined in .

If the response is positive, the RP verifies the correctness of the received Entity Statement, making sure that one of the authority_hints it added to the registration request will lead to the Trust Anchor the OP named using the claim trust_anchor_id. The RP MUST NOT apply metadata policies and assertions from the Trust Chains that the OP provides because those are not valid for the RP's metadata. The RP MUST apply policies and assertions to the metadata using one of its own Trust Chains that ends in the Trust Anchor that the OP chose. Once it has applied those policies and assertions, it can then apply the policies and assertions returned from the OP. This is equivalent to adding the OP's metadata policies and assertions to the Trust Chain in between the RP's and its immediate superior's Entity Statements. When the RP has applied all the metadata policies and assertions to its metadata statement, it then stores the result and will use the agreed-upon metadata when talking to the OP. At this point the RP also knows which Trust Chain it should use when evaluating the OP's metadata. It can therefore apply the metadata policies and assertions on the OP's metadata using the relevant Trust Chain and store the result as the OP's metadata. If the RP does not accept the received Entity Statement for some reason, then it has the choice to restart the registration process or to give up.

A client registration using the explicit method is not expected to be valid forever. The Entity Statements exchanged all have expiration times, which means that the registration will eventually time out. An OP can also, for administrative reasons, decide that a client registration is not valid anymore. An example of this could be the OP leaving the federation used to register an RP. The temporary nature of explicit registration means that an RP must expect its registration to become invalidated at any time, causing RP requests to the OP, such as authorization, token or UserInfo requests, to fail. An RP MAY devise appropriate strategies to re-register with the OP and restart the transaction when such a condition occurs.

At regular intervals, the RP MUST: Starting with the OP's Entity Statement, resolve and verify the Trust Chains it chooses to use when constructing the registration request. If those Trust Chains do not exist anymore or do not verify, then the registration SHOULD be regarded as invalid and a new registration process SHOULD be started. If the OP's Entity Statement was properly formed the RP must now verify that the Entity Statement it received about itself from the OP is still valid. Again, if that is not the case the registration SHOULD be regarded as invalid and a new registration process SHOULD be started. What is regarded as reasonable intervals will depend on federation policies and risk assessment by the maintainer of the RP.

At regular intervals, the OP MUST: If the signature on the registration request has expired, it MUST mark the registration as invalid and demand that the RP MUST re-register. Else starting with the RP's client registration request, the OP MUST verify that there still is a valid Trust Chain terminating in the Trust Anchor the OP chose during the registration process.

An OP MUST NOT assign an expiration time to an RP's registration that is later than the trust chain's expiration time.

The primary differences between Automatic Registration and Explicit Registration are: With Automatic Registration, there is no registration step prior to the Authentication Request, whereas with Explicit Registration, there is. (OpenID Connect Dynamic Client Registration 1.0 and OAuth 2.0 Dynamic Client Registration) also employ a prior registration step.) With Automatic Registration, the Client ID value is the RP's Entity Identifier, and is supplied to the OP by the RP, whereas with Explicit Registration, a Client ID is assigned by the OP and supplied to the RP. Instead of using a Client Secret to authenticate the client, with Automatic Registration, the client is authenticated by means of the RP proving that it controls a private key corresponding to one of its Entity Configuration's public keys.

Both Automatic and Explicit Client Registration support the submission of the Trust Chain embedded in the Request, calculated by the requestor and related to itself. This enables the following benefits in a federation: It solves the problem of OPs using RP metadata that has become stale. This may occur when the OP uses cached RP metadata from a Trust Chain that hasn't reached its expiration time yet. By passing the trust_chain in the Request the RP can notify the OP that a change has taken place, thus letting the OP update its client registration and prevent potential temporary faults due to stale metadata. It enables the RP to pass a verifiable hint which trust path to take in order to build the Trust Chain. This can reduce the costs of RP Federation Entity Discovery for OPs in complex federations where the RP has multiple Trust Anchors or the Trust Chain resolution may result in dead-ends. It enables direct passing of the Entity Configuration, including any present Trust Marks, thus saving the OP from having to make an HTTP request to the RP /.well-known/openid-federation endpoint.

Human-readable Claim Values and Claim Values that reference human-readable values MAY be represented in multiple languages and scripts. This specification enables such representations in the same manner as defined in Section 5.2 of OpenID Connect Core 1.0. The paragraphs that follow are excerpted from there. To specify the languages and scripts, BCP47 language tags are added to member names, delimited by a # character. For example, family_name#ja-Kana-JP expresses the Family Name in Katakana in Japanese, which is commonly used to index and represent the phonetics of the Kanji representation of the same represented as family_name#ja-Hani-JP. As another example, both website and website#de Claim Values might be returned, referencing a Web site in an unspecified language and a Web site in German. Since Claim Names are case sensitive, it is strongly RECOMMENDED that language tag values used in Claim Names be spelled using the character case with which they are registered in the IANA "Language Subtag Registry" . In particular, normally language names are spelled with lowercase characters, region names are spelled with uppercase characters, and scripts are spelled with mixed case characters. However, since BCP47 language tag values are case insensitive, implementations SHOULD interpret the language tag values supplied in a case-insensitive manner. Per the recommendations in BCP47, language tag values for Claims SHOULD only be as specific as necessary. For instance, using fr might be sufficient in many contexts, rather than fr-CA or fr-FR. Where possible, OPs SHOULD try to match requested Claim locales with Claims it has. For instance, if the Client asks for a Claim with a de (German) language tag and the OP has a value tagged with de-CH (Swiss German) and no generic German value, it would be appropriate for the OP to return the Swiss German value to the Client. (This intentionally moves as much of the complexity of language tag matching to the OP as possible, to simplify Clients.)

This specification registers the following metadata names in the IANA "OAuth Authorization Server Metadata" registry established by .

Metadata Name: client_registration_types_supported Metadata Description: Client Registration Types Supported Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): Section of [[ this specification ]] Metadata Name: organization_name Metadata Description: Human-readable name representing the organization owning the OP Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): of [[ this specification ]] Metadata Name: federation_registration_endpoint Metadata Description: Federation Registration Endpoint Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): of [[ this specification ]] Metadata Name: request_authentication_methods_supported Metadata Description: Authentication request authentication methods supported Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): of [[ this specification ]] Metadata Name: request_authentication_signing_alg_values_supported Metadata Description: JSON array containing the JWS signing algorithms supported for the signature on the JWT used to authenticate the request Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): of [[ this specification ]] Metadata Name: signed_jwks_uri Metadata Description: URI pointing to a signed JWT having the Entity's JWK Set as its payload Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): of [[ this specification ]] Metadata Name: jwks Metadata Description: JSON Web Key Set document, passed by value Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): of [[ this specification ]]

This specification registers the following client metadata definition in the IANA "OAuth Dynamic Client Registration Metadata" registry established by .

This specification defines a metadata format that an OAuth 2.0 client can use to obtain the information needed to interact with an OAuth 2.0 protected resource.

Client Metadata Name: client_registration_types Client Metadata Description: Array of strings specifying the client registration types the RP wants to use Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): Section of [[ this specification ]] Client Metadata Name: organization_name Client Metadata Description: Human-readable name representing the organization owning the RP Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): Section of [[ this specification ]] Client Metadata Name: signed_jwks_uri Client Metadata Description: URI pointing to a signed JWT having the Entity's JWK Set as its payload Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): Section of [[ this specification ]]

This section registers the following values in the IANA "OAuth Extensions Error Registry" registry established by .

Name: missing_trust_anchor Usage Location: Authorization Request Protocol Extension: OpenID Connect Federation Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Reference: of [[ this specification ]] Name: validation_failed Usage Location: Authorization Request Protocol Extension: OpenID Connect Federation Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Reference: of [[ this specification ]]

This section registers the following media types in the "Media Types" registry in the manner described in .

Type name: application Subtype name: entity-statement+jwt Required parameters: n/a Optional parameters: n/a Encoding considerations: binary; An Entity Statement is a JWT; JWT values are encoded as a series of base64url-encoded values (some of which may be the empty string) separated by period ('.') characters. Security considerations: See of [[ this specification ]] Interoperability considerations: n/a Published specification: [[ this specification ]] Applications that use this media type: Applications that use [[ this specification ]] Fragment identifier considerations: n/a Additional information: Magic number(s): n/a File extension(s): n/a Macintosh file type code(s): n/a Person & email address to contact for further information: Michael B. Jones, mbj@microsoft.com Intended usage: COMMON Restrictions on usage: none Author: Michael B. Jones, mbj@microsoft.com Change controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Provisional registration? No Type name: application Subtype name: trust-mark+jwt Required parameters: n/a Optional parameters: n/a Encoding considerations: binary; A Trust Mark is a JWT; JWT values are encoded as a series of base64url-encoded values (some of which may be the empty string) separated by period ('.') characters. Security considerations: See of [[ this specification ]] Interoperability considerations: n/a Published specification: [[ this specification ]] Applications that use this media type: Applications that use [[ this specification ]] Fragment identifier considerations: n/a Additional information: Magic number(s): n/a File extension(s): n/a Macintosh file type code(s): n/a Person & email address to contact for further information: Michael B. Jones, mbj@microsoft.com Intended usage: COMMON Restrictions on usage: none Author: Michael B. Jones, mbj@microsoft.com Change controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Provisional registration? No Type name: application Subtype name: resolve-response+jwt Required parameters: n/a Optional parameters: n/a Encoding considerations: binary; An Entity Resolve Response is a signed JWT; JWT values are encoded as a series of base64url-encoded values (some of which may be the empty string) separated by period ('.') characters. Security considerations: See of [[ this specification ]] Interoperability considerations: n/a Published specification: [[ this specification ]] Applications that use this media type: Applications that use [[ this specification ]] Fragment identifier considerations: n/a Additional information: Magic number(s): n/a File extension(s): n/a Macintosh file type code(s): n/a Person & email address to contact for further information: Michael B. Jones, mbj@microsoft.com Intended usage: COMMON Restrictions on usage: none Author: Michael B. Jones, mbj@microsoft.com Change controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Provisional registration? No Type name: application Subtype name: trust-chain+json Required parameters: n/a Optional parameters: n/a Encoding considerations: binary; A Trust Chain is a JSON Array of JWTs; JWT values are encoded as a series of base64url-encoded values (some of which may be the empty string) separated by period ('.') characters. Security considerations: See of [[ this specification ]] Interoperability considerations: n/a Published specification: [[ this specification ]] Applications that use this media type: Applications that use [[ this specification ]] Fragment identifier considerations: n/a Additional information: Magic number(s): n/a File extension(s): n/a Macintosh file type code(s): n/a Person & email address to contact for further information: Michael B. Jones, mbj@microsoft.com Intended usage: COMMON Restrictions on usage: none Author: Michael B. Jones, mbj@microsoft.com Change controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Provisional registration? No

This specification registers the following parameter name in the IANA "OAuth Parameters" registry established by .

Parameter Name: trust_chain Parameter Usage Location: authorization request Change Controller: OpenID Foundation Artifact Binding Working Group - openid-specs-ab@lists.openid.net Specification Document(s): Section of [[ this specification ]]

Some of the interfaces defined in this specification could be used for Denial-of-Service attacks (DoS), most notably, the Resolve endpoint (), Explicit Client Registration (), and Automatic Client Registration (), can be exploited as vectors of http propagation attacks. If these endpoints are provided, some adequate defense methods are required, such as those described below and in . A Trust Mark can be statically validated, using the public key of its issuer. The static validation of the Trust Marks represent a filter against a propagation attacks. An attacker could exploit the Federation Entity Discovery mechanism and use an OIDC Federation to propagate many http requests. For each authorization request, crafted by an anonymous client, the OP would produce ~3 http requests to third parties in absence of intermediaries, and at least 5 http requests with at least one intermediary. If an OP doesn't find at least one valid Trust Mark in an Entity Configuration it should reject the request and temporary bans the requestor. If client authentication is not demanded at the Resolve endpoint then incoming requests should not by default result in immediate collection (Federation Entity Discovery process) and evaluation of Trust Chains.

One of the fundamental design goals of this protocol is to protect messages end-to-end. This can not be accomplished by demanding TLS since TLS, in lots of cases, is not end-to-end but ends in a HTTPS to HTTP Reverse Proxy. Allowing unsigned error messages therefore opens up an attack vector for someone who wants to run a Denial of Service attack. This is not specific to OpenID Connect Federation but equally valid for other protocols when HTTPS to HTTP reverse proxies are used.

Nomura Research Institute, Ltd. Ping Identity Microsoft Google Salesforce Nomura Research Institute, Ltd. Ping Identity Microsoft Illumila Nomura Research Institute, Ltd. Ping Identity Microsoft IANA IANA IANA

Let us assume the following: The project LIGO would like to offer access to its wiki to all OPs in EduGAIN. LIGO is registered to the InCommon federation.

The players

Both SWAMID and InCommon are identity federations in their own right. They also have in common that they are both members of the EduGAIN federation. SWAMID and InCommon are different in how they register entities. SWAMID registers organizations and lets the organizations register entities that belong to the organization, while InCommon registers all entities directly and not beneath any organization Entity. Hence the differences in depth in the federations. Let us assume a researcher from Umeå University would like to login at the LIGO Wiki. At the Wiki, the researcher will use some kind of discovery service to find the home identity provider (op.umu.se) Once the RP-part of the Wiki knows which OP it SHOULD talk to it has to find out a couple of things about the OP. All of those things can be found in the metadata. But finding the metadata is not enough; the RP also has to trust the metadata. Let us make a detour and start with what it takes to build a federation.

These are the steps to set up a federation infrastructure. Generation of signing keys. These must be public/private key pairs. Set up a signing service that can sign JWTs/Entity Statements using the Federation Entity Keys. Set up web services that can publish signed Entity Statements, one for the URL corresponding to the federation's Entity Identifier returning an Entity Configuration and the other one providing the fetch Entity Statement endpoint, as described in . Once the previous requirements have been satisfied, a Federation Operator can add Entities to the federation. Adding an Entity comes down to: Providing the Entity with the federation's Entity Identifier and the public part of the key pairs used by the federation operator for signing Entity Statements. Getting the Entity's Entity Identifier and the JWKS that the Entity plans to publish in its Entity Configuration. Now before the federation operator starts adding entities, there have to be policies in place on who can be part of the federation and the layout of the federation. Is it supposed to be a one-layer federation like InCommon, a two-layer one like the SWAMID federation, or a multi-layer federation? The federation may also want to think about implementing other policies using the federation policy framework, as described in . With the federation in place, things can start happening.

Federation Entity Discovery is a sequence of steps that starts with the RP fetching the Entity Configuration of the Leaf Entity (in this case https://op.umu.se) using the process defined in . What follows thereafter is this sequence of steps: Pick out the immediate superior entities using the authority hints Fetch the configuration for each such Entity. This uses the process defined in Using the fetch endpoint of the superiors to ask for information about the subordinate Entity . How many times this has to be repeated depends on the depth of the federation. What follows below is the result of each step the RP has to take to find the OP's metadata using the federation setup described above. When building the Trust Chain, the Entity Statements issued by a superior about its subordinate are used together with the Entity Configuration issued by the Leaf. The Entity Configuration concerning intermediates is not part of the Trust Chain.

The LIGO WIKI RP fetches the Entity Configuration from the OP (op.umu.se) using the process defined in .

The result is this Entity Configuration.

The authority_hints points to the intermediate https://umu.se. So that is the next step. This Entity Configuration is the first link in the Trust Chain.

The LIGO RP fetches the Entity Configuration from "https://umu.se" using the process defined in . The request will look like this:

And the GET will return:

The only piece of information that is used from this Entity Statement is the federation_fetch_endpoint, which is used in the next step.

The RP uses the fetch endpoint provided by https://umu.se as defined in to fetch information about "https://op.umu.se". The request will look like this:

and the result is this:

This is the second link in the Trust Chain. Notable here is that this path leads to two Trust Anchors using the same next step ("https://swamid.se").

The LIGO Wiki RP fetches the Entity Configuration from "https://swamid.se" using the process defined in . The request will look like this:

And the GET will return:

The only piece of information that is used from this Entity Statement is the federation_fetch_endpoint, which is used in the next step.

The LIGO Wiki RP uses the fetch endpoint provided by "https://swamid.se" as defined in to fetch information about "https://umu.se". The request will look like this:

and the result is this:

This is the third link in the Trust Chain. If we assume that the issuer of this Entity Statement is not in the list of Trust Anchors the LIGO Wiki RP has access to, we have to go one step further.

RP fetches the Entity Configuration from "https://edugain.geant.org" using the process defined in . The request will look like this:

And the GET will return:

Again, the only thing we need is the federation_fetch_endpoint. As described in , note SHOULD also be taken to jwks as the Trust Anchor MAY be performing a key rollover.

The LIGO Wiki RP uses the fetch endpoint of https://edugain.geant.org as defined in to fetch information about "https://swamid.se". The request will look like this:

and the result is this:

If we assume that the issuer of this statement appears in the list of Trust Anchors the LIGO Wiki RP has access to this would be the fourth and final Entity Statement in the Trust Chain. We now have the whole chain from the Entity Configuration of the Leaf Entity up until the last one that is issued by a Trust Anchor. All in all, we have: Entity Configuration by the Leaf Entity (https://op.umu.se) Entity Configuration by https://umu.se Statement issued by https://umu.se about https://op.umu.se Entity Configuration by https://swamid.se Statement issued by https://swamid.se about https://umu.se Entity Configuration by https://edugain.geant.org Statement issued by https://edugain.geant.org about https://swamid.se Using the public keys of the Trust Anchor that the LIGO Wiki RP has been provided with in some secure out-of-band way, it can now verify the Trust Chain as described in .

Having verified the chain, the LIGO Wiki RP can proceed with the next step. Combining the metadata policies from the tree Entity Statements we have by a superior about its subordinate and applying the combined policy to the metadata statement that the Leaf Entity presented, we get:

We have now reached the end of the Provider Discovery process.

As described in , there are two ways which can be used to do client registration: No negotiation between the RP and the OP is made regarding what features the client SHOULD use in future communication are done. The RP's published metadata filtered by the chosen trust chain's metadata policies defines the metadata that is to be used. The RP will access the federation_registration_endpoint, which provides the metadata for the RP to use. The OP MAY return a metadata policy that adds restrictions over and above what the Trust Chain already has defined.

The LIGO Wiki RP does not do any registration but goes directly to sending an Authentication Request. Here is an example of such an Authentication Request:

The OP receiving this Authentication Request will, unless the RP is already registered, start to dynamically fetch and establish trust with the RP.

The OP needs to establish a Trust Chain for the RP (wiki.ligo.org). The OP in this example is configured with public keys of two federations: https://edugain.geant.org https://swamid.se The OP starts to resolve metadata for the client identifier https://wiki.ligo.org by fetching the Entity Configuration using the process described in . The process is the same as described in and will result in a Trust Chain with the following Entity Statements: Entity Configuration by the Leaf Entity https://wiki.ligo.org Statement issued by https://incommon.org about https://wiki.ligo.org Statement issued by https://edugain.geant.org about https://incommon.org

Using the public keys of the Trust Anchor that the LIGO Wiki RP has been provided with in some secure out-of-band way, it can now verify the Trust Chain as described in . We will not list the complete Entity Statements but only the metadata and metadata_policy parts. There are two metadata policies:

Combining these and apply them to the metadata for wiki.ligo.org:

The final result is:

Having that, the registration is done, and the OP MUST now use the keys found at the URL specified in signed_jwks_uri to verify the signature on the Request Object in the Authentication Request.

Here the LIGO Wiki RP sends a client registration request to the federation_registration_endpoint of the OP (op.umu.se). What it sends is an Entity Configuration. Once the OP has the Entity Statement, it proceeds with the same sequence of steps as laid out in . The OP will end up with the same RP metadata as was described in , but what it now can do is return a metadata policy that it wants to be applied to the RP's metadata. This metadata policy will be combined with the Trust Chain's combined metadata policy before being applied to the RP's metadata. If we assume that the OP does not support refresh tokens, it MAY want to add a metadata policy that says:

Thus, the Entity Statement returned by the OP to the RP MAY look like this:

And the resulting metadata used by the RP could look like:

Copyright (c) 2022 The OpenID Foundation. The OpenID Foundation (OIDF) grants to any Contributor, developer, implementer, or other interested party a non-exclusive, royalty free, worldwide copyright license to reproduce, prepare derivative works from, distribute, perform and display, this Implementers Draft or Final Specification solely for the purposes of (i) developing specifications, and (ii) implementing Implementers Drafts and Final Specifications based on such documents, provided that attribution be made to the OIDF as the source of the material, but that such attribution does not indicate an endorsement by the OIDF. The technology described in this specification was made available from contributions from various sources, including members of the OpenID Foundation and others. Although the OpenID Foundation has taken steps to help ensure that the technology is available for distribution, it takes no position regarding the validity or scope of any intellectual property or other rights that might be claimed to pertain to the implementation or use of the technology described in this specification or the extent to which any license under such rights might or might not be available; neither does it represent that it has made any independent effort to identify any such rights. The OpenID Foundation and the contributors to this specification make no (and hereby expressly disclaim any) warranties (express, implied, or otherwise), including implied warranties of merchantability, non-infringement, fitness for a particular purpose, or title, related to this specification, and the entire risk as to implementing this specification is assumed by the implementer. The OpenID Intellectual Property Rights policy requires contributors to offer a patent promise not to assert certain patent claims against other contributors and against implementers. The OpenID Foundation invites any interested party to bring to its attention any copyrights, patents, patent applications, or other proprietary rights that MAY cover technology that MAY be required to practice this specification.

The authors wish to acknowledge the contributions of the following individuals and organizations to this specification: Heather Flanagan, Samuel Gulliksson, Takahiko Kawasaki, Torsten Lodderstedt, Roberto Polli, Jouke Roorda, Mischa Sallé, Marcos Sanz, Peter Schober, Michael Schwartz, Kristina Yasuda, and the JRA3T3 task force of GEANT4-2.

[[ To be removed from the final specification ]] -25 Fixed #1684: disambiguation on the role federation_entity. Fixed #1703: Unsigned error response Fixed #1693: Trust mark status endpoint HTTP method changed to POST. Fixed #1681: RS256 is not mandatory anymore for federation operations signatures. Fixed #1701: trust_chain parameter with PAR and redirect_uri. Fixed #1695: entity_type url paramenter for Listing endpoint. Fixed #1712: federation_entity claims: organization_name is recommended, logo_uri is added. Fixed #1702: removal of metadata type trust_mark_issuer, federation_status_endpoint moved in the federation_entity metadata and renamed to federation_trust_mark_status_endpoint. Fixed #1716: Clarified how to retrieve the RP's Entity Configuration when using Automatic Registration. Fixed #1656: Introductionary text review. -24 Fixed #1654: Text cleanup on how and who publishes the entity configuration. Relaxed JWK Set representations: jwks, signed_jwks_uri and jwks_uri can coexist in the same Metadata for interoperability purposes across different Federations. Fixed #1641: Federation Historical Keys endpoint. Fixed #1673: added resolve endpoint JWT claims. Fixed #1663: clarifications on the http status codes returning from /.well-known/openid-federation. Fixed #1667: serialization format for federation endpoint made explicit. Fixed #1662: ID Token section removed. Fixed #1680: removal of the claim operation in Generic Errors. Fixed #1669: error types in the section Generic Error Response. Added Vladimir Dzhuvinov as an editor. -23 Text on max_path_length Fixed #1634: Trust Chain explanatory text. Federation Entity Discovery as defined term. Fixed #1645: Federation Entity Keys as defined term. Fixed #1633: Explaination text about who signs the Trust Mark and how. Fixed #1650: typo in signed_jws_uri and jwks metadata claims. -22 Rewritten text about metadata processing when doing Explicit Client Registration. Fixed #1581: OP metadata claims enabled for AS. Fixed #1594: self-issued trust_chain in the Authorization Request. Fixed #1583: metadata policy one_of operator, explanatory text for multiple values. Fixed #1584: Stated that domain name constraints are as specified in Section 4.2.1.10 of . Added more text on considerations when using the resolve endpoint. Removal of aud parameter in the fetch endpoint request. General rewording with the terms Leaf Entity and Entity Configuration. Fixed #1588: Explicitly described the differences between Automatic and Explicit Registration. Fixed #1606: Described situation in which the requirement for signed requests with Automatic Registration could be relaxed. Fixed #1629: authz request object, audience explanatory text. Fixed #1630: Resolve endpoint response content type. Fixed #1608: submission of the Trust Chain in the Explicit Client Registration Request. -21 Fixed #1547: Metadata section restructured. Fixed #1548: request_authentication_signing_alg_values_supported clarification text about the usage of private_key_jwt and request_object. Added a new constraint, allowed_leaf_entity_types. Resolve endpoint: Removed iss paramenter in the request and specified the usage of the aud claim in the response. Fixed #1513: request_authentication_methods_supported according to IANA OAuth2 AS metadata registered names. Fixed #1527: Defined the Trust Mark term and added non-normative examples. Added Security Considerations regarding the role of the Trust Marks. Editorial review, normative language, and typos. Fixed #1535: Removed the sub claim from the non-normative example of the Authorization Request Object. Fixed #1528: Added Claims Languages and Scripts section. Fixed #1456: Added language about space-delimited string parameters from RFC 6749. -20 Updated example for Resolve endpoint. Recommended that Key IDs be the JWK Thumbprint of the key. Fixed #1502 - Corrected usage of id_token_signed_response_alg. Fixed #1506 - Referenced RFC 9126 for Pushed Authorization Requests. Added Giuseppe De Marco as an editor and removed Samuel Gulliksson as an editor. Fixed #1508 - Editorial review. Fixed #1505 - Defined that signed_jwks_uri is preferred over jwks_uri and jwks. Fixed #1514 - Corrected RP metadata examples to use id_token_signed_response_alg. Fixed #1512 - Registered the error codes missing_trust_anchor and validation_failed. Also registered the OP Metadata parameters organization_name, request_authentication_methods_supported, request_authentication_signing_alg_values_supported, signed_jwks_uri, and jwks and the RP Metadata parameters client_registration_types (which was previously misregistered as client_registration_type), organization_name, and signed_jwks_uri. Defined the term Federation Operator and described redundant retrieval of Trust Anchor keys. Fixed #1519 - Corrected instances of client_registration_type to client_registration_types_supported. Fixed #1520 - Renamed federation_api_endpoint to federation_fetch_endpoint. Fixed #1521 - Changed swamid.sunet.se to swamid.se in examples. Fixed #1523 - Added request_parameter_supported to examples. Capitalized defined terms. -19 Fixed #1497 - Removed trust_mark claim from federation entity metadata. Fixed #1446 - Added trust_chain and removed is_leaf. Fixed #1479 - Added jwks claim in OP metadata. Fixed #1477 - Added request_authentication_methods_supported. Fixed #1474 - Added the request_authentication_signing_alg_values_supported OP metadata parameter. Rewrote trust_marks definition. -18 Added the jwks claim name for OP Metadata. Moved from a federation API to separate endpoints per operation. Added descriptions of the list, resolve and status federation endpoints. Registered the client_registration_types_supported and federation_registration_endpoint authorization server metadata parameters. Registered the client_registration_type dynamic client registration parameter. Registered and used the application/entity-statement+jwt media type. Registered the application/trust-mark+jwt media type. Trust marks: the claim mark has been renamed to logo_uri. New federation endpoint: Resolve Entity Statement. New federation endpoint: Trust Mark Status. Federation API, Fetch: iss parameter is optional. Trust Marks: added non-normative examples. Expanded Section 8.1: Included entity configurations. Fixed #1373 - More clearly defined the term Entity Statement, both in the Terminology section and in the Entity Statement section. -17 Addressed many working group review comments. Changes included adding the Overall Architecture section, the trust_marks_issuers claim, and the Setting Up a Federation section. -16 Added Security Considerations section on Denial-of-Service attacks. -15 Added signed_jwks_uri, which had been lost somewhere along the road. -14 Rewrote the federation policy section. What previously was referred to as client authentication in connection with automatic client registration is now changed to be request authentication. Made a distinction between parameters and claims. Corrected the description of the intended behavior when essential is absent. -13 Added 'trust_marks' as a parameter in the entity statement. An RP's self-signed entity statement MUST have the OP's issuer identifier as the value of 'aud'. Added RS256 as default signing algorithm for entity statements. Specified that the value of 'aud' in the entity statement use in automatic client registration MUST have the ASs authorization endpoint URL as value. Also, the 'sub' claim MUST NOT be present. Separating the usage of merge and combine as proposed by Vladimir Dzhuvinov in Bitbucket issue #1157. An RP doing only explicit client registration are not required to publish and support a /.well-known/openid-federation. Every key in a JWK Set in an entity statement MUST have a kid. -12 Made JWK Set OPTIONAL in an entity statement returned by OP as response of an explicit client registration Renamed entity type to client registration type. Added more text describing client_registration_auth_methods_supported. Made "Processing the Authentication Request" into two separate sections: one for Authentication Request and one for Pushed Authorization Request. Added example of URLs to some examples in the appendix. Changed the automatic client registration example in the appendix to use Request Object instead of a client_assertion. -11 Added section on trust marks. Clarified private_key_jwt usage in the authentication request. Fixed Bitbucket issues #1150 and #1155 by Vladimir Dzhuvinov. Fixed some examples to make them syntactically correct. -10 Incorporated additional review feedback from Marcos Sanz. The primary change was moving constraints to their own section of the entity statement. -09 Incorporated review feedback from Marcos Sanz. Major changes were as follows. Separated entity configuration discovery from operations provided by the federation API. Defined new authentication error codes. Also incorporated review feedback from Michael B. Jones. -08 Incorporated review feedback from Michael B. Jones. Major changes were as follows. Deleted sub_is_leaf entity statement since it was redundant. Added client_registration_type RP registration metadata value and client_registration_types_supported OP metadata value. Deleted openid_discovery metadata type identifier since its purpose is already served by openid_provider. Entity identifier paths are now included when using the Federation API, enabling use in multi-tenant deployments sharing a common domain name. Renamed sub_is_leaf to is_leaf in the Entity Listings Request operation parameters. Added crit and policy_language_crit, enabling control over which entity statement and policy language extensions MUST be understood and processed. Renamed openid_client to openid_relying_party. Renamed oauth_service to oauth_authorization_server. Renamed implicit registration to automatic registration to avoid naming confusion with the implicit grant type. Renamed op to operation to avoid naming confusion with the use of "OP" as an acronym for "OpenID Provider". Renamed url to uri in several identifiers. Restored Open Issues appendix. Corrected document formatting issues. -07 Split metadata into metadata and metadata_policy Updated example -06 Some rewrites Added example of explicit client registration -05 A major rewrite. -04 Changed client metadata names scopes to rp_scopes and claims to rp_claims. Added Open Issues appendix. Added additional references. Editorial improvements. Added standard Notices section, which is present in all OpenID specifications.