Network Working Group Y. Wang
Internet-Draft HJS Foundation
Intended status: Informational Ltd.
Expires: 20 September 2026 21 March 2026
Co-Evolve Binding Protocol-00 (CEP)
draft-wang-cep-00
Abstract
This document specifies the Co-Evolve Binding Protocol (CEP), a
lightweight, minimal-structured protocol designed to bind human
cognitive evolution with artificial intelligence (AI) evolution.
CEP works in conjunction with the Judgment Event Protocol (JEP) and
Human Judgment Structure (HJS) as the default collaborative system,
while supporting adaptation to other compliant judgment and
accountability systems to enhance interoperability and practicality.
Leveraging JEP's strengths in standardized judgment recording and
lightweight verifiability, HJS's advantages in structured human
responsibility management and end-to-end accountability traceability,
and the flexibility to adapt to other systems, CEP integrates core
advantages to form a complementary three-layer human-AI symbiosis
infrastructure. The core goal of CEP is to prevent unilateral AI
evolution, maintain human sovereignty in the co-evolution process of
humans and AI, and lay the foundation for trustworthy human-AI
symbiosis.
Status of This Memo
This Internet-Draft is submitted in full conformance with the
provisions of BCP 78 and BCP 79.
Internet-Drafts are working documents of the Internet Engineering
Task Force (IETF). Note that other groups may also distribute
working documents as Internet-Drafts. The list of current Internet-
Drafts is at https://datatracker.ietf.org/drafts/current/.
Internet-Drafts are draft documents valid for a maximum of six months
and may be updated, replaced, or obsoleted by other documents at any
time. It is inappropriate to use Internet-Drafts as reference
material or to cite them other than as "work in progress."
This Internet-Draft will expire on 20 September 2026.
Copyright Notice
Copyright (c) 2026 IETF Trust and the persons identified as the
document authors. All rights reserved.
This document is subject to BCP 78 and the IETF Trust's Legal
Provisions Relating to IETF Documents
(https://trustee.ietf.org/license-info) in effect on the date of
publication of this document. Please review these documents
carefully, as they describe your rights and restrictions with respect
to this document.
Table of Contents
1. Introduction .................................................... 3
1.1. Background ................................................ 3
1.2. Objectives ................................................ 4
1.3. Scope ..................................................... 5
2. Terminology ..................................................... 6
3. Protocol Overview ............................................... 7
3.1. Architecture Positioning .................................. 7
3.2. Core Design Principles .................................... 8
4. Protocol Specification ......................................... 10
4.1. Core Fields (Mandatory) .................................. 10
4.1.1. HumanAnchor ......................................... 11
4.1.2. ShiftClass .......................................... 11
4.1.3. BoundState .......................................... 12
4.2. Optional Extended Fields ................................. 13
4.3. Protocol Frame Format .................................... 14
4.4. Protocol Interaction Process ............................. 15
4.5. Adaptation Specifications for Other Compliant Systems .... 16
5. Security Considerations ....................................... 18
6. Privacy Considerations ........................................ 18
7. IANA Considerations ........................................... 19
8. References .................................................... 19
8.1. Normative References ..................................... 19
8.2. Informative References ................................... 20
Author's Address .................................................. 20
1. Introduction
1.1. Background
With the rapid development of AI technology, especially the emergence
of large language models and autonomous agents, AI is increasingly
moving towards self-evolution, self-optimization, and closed-loop
reasoning. This trend brings the risk of AI divergence from human
cognitive logic, value systems, and responsibility boundaries—
leading to a "civilization bifurcation" between human civilization
and silicon-based intelligence.
To address this risk, the Judgment Event Protocol (JEP)
[I-D.wang-jep-judgment-event-protocol] provides a minimal, verifiable
format for logging judgment-related actions, ensuring standardized
human participation in key AI decision-making processes with its
lightweight design and replay protection mechanism. The Human
Judgment Structure (HJS) [I-D.wang-hjs-judgment-structure], as an
infrastructure-grade accountability layer built on JEP, complements
JEP's advantages by providing a structured framework for human
responsibility management, cryptographic accountability mechanisms,
and a three-tier privacy architecture, ensuring traceability of
causality and accountability in AI operations. However, there is a
lack of a dedicated protocol to bind human and AI evolution, leading
to potential unilateral AI evolution and loss of human sovereignty in
the evolution process—CEP is designed to fill this gap by integrating
JEP and HJS's core strengths, while supporting adaptation to other
compliant judgment and accountability systems to meet diverse
practical deployment needs.
1.2. Objectives
CEP is designed to achieve the following core objectives, aligning
with JEP, HJS and other compliant systems, and integrating their
strengths to form a three-layer human-AI symbiosis infrastructure
with strong adaptability:
1. Bind AI evolution to human cognitive evolution, preventing
unilateral AI evolution and civilization bifurcation, leveraging
JEP's standardized judgment records, HJS's structured
accountability, and the adaptability to other systems to ensure
traceable and controllable co-evolution.
2. Provide minimal-structured, extensible fields to cover all AI
evolution scenarios (e.g., fine-tuning, emergence, drift, loop
formation), inheriting JEP's minimal design concept, HJS's
extensible architecture, and supporting flexible adaptation to
other systems without modifying the core structure.
3. Ensure seamless compatibility with JEP, HJS, and other compliant
judgment/accountability systems, integrating their verification
mechanisms and accountability traceability capabilities to enable
end-to-end human-AI co-evolution, decision-making, and
accountability.
4. Be lightweight, easy to implement, and compatible with existing
AI systems, network protocols, and W3C standards, adhering to
JEP's lightweight engineering design, HJS's infrastructure-grade
compatibility requirements, and the practical deployment needs of
other systems.
1.3. Scope
CEP applies to all AI systems, large models, and intelligent agents
that involve self-evolution, self-optimization, or capability
emergence. It focuses on the binding of human and AI evolution
processes, without involving AI model training, algorithm design, or
hardware implementation. Drawing on JEP's non-intrusive design,
HJS's middleware-style deployment, and the flexible adaptation
capability to other systems, CEP works as a middleware layer,
inserting into the evolution process of AI systems to ensure human
oversight and symbiotic evolution, while maintaining compatibility
with JEP's judgment recording, HJS's accountability traceability, and
the core functions of other compliant systems.
2. Terminology
The following terms are used throughout this document:
CEP
Co-Evolve Binding Protocol, the protocol specified in this
document, which integrates the strengths of JEP, HJS and other
compliant systems to bind human and AI evolution.
JEP
Judgment Event Protocol [I-D.wang-jep-judgment-event-protocol],
a minimal, verifiable format for logging judgment-related
actions in distributed systems, which defines four immutable
event verbs (J, D, T, V), a signed JSON event structure with
replay protection, and standardized signature verification
rules. Its core strengths lie in lightweight design,
standardized judgment recording, and reliable verification,
serving as the default judgment system for CEP and providing
structured human judgment records and a verification interface.
HJS
Human Judgment Structure [I-D.wang-hjs-judgment-structure], an
infrastructure-grade accountability layer for AI agents built
on the Judgment Event Protocol (JEP), which inherits JEP's four
core primitives (Judge, Delegate, Terminate, Verify) and
complements JEP's advantages with a structured human
responsibility management framework, a cryptographic mechanism
for cross-platform responsibility verification, a minimal
receipt structure with responsibility chaining, and a three-
tier privacy architecture for accountability governance. It
serves as the default accountability system for CEP, fully
compatible with CEP's human-AI co-evolution binding logic,
providing end-to-end accountability traceability.
Compliant Judgment/Accountability Systems
Other judgment or accountability systems that meet international
standards (e.g., RFC, W3C) and core requirements of human-AI
symbiosis (e.g., human sovereignty, traceability,
accountability), which CEP can adapt to on the premise of not
changing its core structure.
HumanAnchor
A core field in CEP, used to bind AI evolution to a specific
human judgment or value anchor (associated with the judgment ID
of JEP or other compliant judgment systems), leveraging the
standardized judgment recording of JEP/other systems and the
responsibility binding of HJS/other compliant accountability
systems to ensure traceability of the binding relationship.
ShiftClass
A core field in CEP, used to classify the type of AI evolution
(e.g., fine-tuning, emergence, drift), enabling targeted
supervision and aligning with the structured accountability
management requirements of HJS and other compliant
accountability systems.
BoundState
A core field in CEP, used to mark whether AI evolution is bound
to human cognitive evolution (enabling symbiotic evolution),
serving as a "switch" for co-evolution and integrating the
responsibility closure mechanism of HJS and other compliant
accountability systems to ensure that unbound unilateral
evolution is rejected.
Co-Evolution
The process in which AI and humans evolve synchronously, with
AI's evolution amplifying human cognitive capabilities and
human's evolution constraining AI's boundary, supported by the
judgment participation of JEP/other compliant systems and the
accountability traceability of HJS/other compliant systems.
Unilateral Evolution
The process in which AI evolves independently without human
oversight, value alignment, or responsibility binding, which
CEP prevents by integrating the joint constraints of JEP, HJS,
and other compliant systems.
3. Protocol Overview
3.1. Architecture Positioning
CEP is the third layer of the human-AI symbiosis infrastructure,
working with JEP, HJS, and other compliant judgment/accountability
systems, and integrating their core strengths to form a three-layer
closed-loop system with strong adaptability. Each layer complements
the others, with JEP/other compliant judgment systems providing
standardized judgment support, HJS/other compliant accountability
systems offering structured accountability protection, and CEP
realizing co-evolution binding, forming a complete human-AI symbiosis
system:
Bottom Layer: HJS and other compliant accountability systems. HJS, as
the default infrastructure-grade accountability layer for AI
agents built on JEP, inherits JEP's lightweight and verifiable
advantages and complements them with a structured human
responsibility management framework. Other compliant
accountability systems that meet core requirements can also be
adapted, providing a cryptographic mechanism to manage and verify
responsibility for AI agent judgment behaviors across
heterogeneous platforms, defining core primitives compatible with
CEP, with a minimal receipt structure and privacy architecture for
accountability governance. They enable traceability of AI
evolution events, complete the closed-loop of human-AI co-
evolution accountability, and provide reliable accountability
support for CEP.
Middle Layer: JEP and other compliant judgment systems. JEP, with its
core strengths of lightweight design, standardized judgment
recording, and reliable verification, ensures human participation
in key decision-making, serves as the default judgment system for
CEP, and provides structured human judgment records and a
verification interface. Other compliant judgment systems that meet
core requirements can also be adapted, providing standardized
human judgment records compatible with CEP's HumanAnchor field,
laying the foundation for CEP's co-evolution binding.
Top Layer: CEP (Co-Evolve Binding Protocol). Integrates the
standardized judgment verification of JEP/other compliant judgment
systems and the structured accountability traceability advantages
of HJS/other compliant accountability systems, binds human and AI
evolution, locks the "non-divergence" of human and AI
civilizations, and realizes the organic connection between
judgment participation and accountability closure of various
compliant systems.
3.2. Core Design Principles
CEP adheres to the following design principles, consistent with
IETF's preference for lightweight, engineering-oriented, and
standardized protocols, while fully integrating the core strengths of
JEP, HJS, and other compliant systems:
Minimal Structuring: Adopt a minimal-structured design consistent
with JEP's lightweight concept, HJS's minimal receipt structure,
and the practical needs of other compliant systems, using only 3
core mandatory fields to avoid redundant design, reduce the cost
of protocol implementation and maintenance, and align with the
cost optimization requirements of all compliant systems, forming a
complementary relationship with their cryptographic erasure and
privacy architecture to achieve cost structure optimization.
Interoperability: Be compatible with existing AI systems, network
protocols, JEP, HJS, and other compliant judgment/accountability
systems, leveraging their standardized formats and cross-platform
compatibility to enable seamless integration and global
interoperability.
Engineering-Oriented: Use mature IETF terminology (e.g., Binding,
State) to avoid abstract or biological terms, following the
engineering-oriented design concept of JEP, HJS, and other
compliant systems, ensuring readability and implementability by
engineers.
Extensibility and Adaptability: Support optional extended fields to
adapt to future AI evolution scenarios and the characteristics of
other compliant systems without changing the core structure,
inheriting the extensible architecture of HJS, the flexible
verification mechanism of JEP, and the adaptation needs of other
systems.
Human Sovereignty: Ensure that human beings have the final right to
control AI evolution, and AI evolution cannot be separated from
human oversight, integrating the human judgment participation of
JEP/other compliant judgment systems and the human responsibility
management advantages of HJS/other compliant accountability
systems to safeguard human sovereignty in co-evolution.
4. Protocol Specification
4.1. Core Fields (Mandatory)
CEP defines 3 mandatory core atomic fields, which are minimal,
structured, and compatible with JEP, HJS, and other compliant
judgment/accountability systems. All AI evolution events must carry
these 3 fields to ensure symbiotic binding. The design of these
fields fully integrates the standardized advantages of JEP/other
compliant judgment systems and the accountability requirements of
HJS/other compliant accountability systems, ensuring lightweight
implementation while maintaining traceability and controllability.
4.1.1. HumanAnchor
Definition: The unique identifier of the human judgment or value
anchor bound to AI evolution, directly associated with the
judgment event ID (or equivalent core identifier) in JEP or other
compliant judgment systems. This field integrates the standardized
judgment recording advantage of JEP/other compliant judgment
systems and the responsibility binding requirement of HJS/other
compliant accountability systems, ensuring that AI evolution is
anchored to human cognitive logic and value boundaries, and the
validity of this identifier is verified through the standardized
signature verification rules of JEP/other compliant judgment
systems, while the binding relationship is traceable through the
accountability system of HJS/other compliant accountability
systems.
Format: Consistent with the judgment event identifier format of JEP
or other compliant judgment systems (for JEP, this is the UUIDv4
nonce field or the multihash "what" field of JEP events; for other
compliant systems, as defined in their core specifications),
ensuring interoperability between CEP and JEP/other compliant
judgment systems, and compatible with the accountability log ID
format of HJS/other compliant accountability systems to facilitate
traceability.
Constraints: Cannot be empty; must correspond to a valid judgment
event in JEP or other compliant judgment systems (verified via the
signature verification steps of the corresponding system,
including nonce uniqueness, timestamp validity, and signature
correctness as defined in Section 2.4 of [I-D.wang-jep-judgment-
event-protocol]), and the associated human responsibility
information must be recorded in HJS or other compliant
accountability systems to ensure accountability closure. If the
associated judgment event is invalid or the responsibility record
is missing, the AI evolution event is rejected.
4.1.2. ShiftClass
Definition: Classifies the type of AI evolution, used to identify the
nature of AI evolution and implement targeted supervision. This
field is designed in line with the structured accountability
management advantage of HJS and other compliant accountability
systems, enabling them to perform classified accountability based
on evolution types, while inheriting the standardized enumeration
design of JEP and other compliant systems to ensure consistency.
Enumeration Values (fixed, non-extensible to ensure standardization,
consistent with the standardized design concept of JEP, HJS, and
other compliant systems):
FINE_TUNE: AI model parameter fine-tuning, incremental evolution.
EMERGENCE: Sudden emergence of new capabilities that were not
explicitly trained.
DRIFT: Deviation of AI's values, preferences, or reasoning logic
from the original human alignment.
LEAP: Qualitative leap in AI capabilities (e.g., from narrow AI to
general AI in a specific field).
LOOP: AI forms a closed-loop reasoning chain that is independent
of human oversight.
Constraints: Must be one of the above enumeration values; no custom
values are allowed, ensuring standardization consistent with JEP,
HJS, and other compliant systems, and facilitating unified
supervision and accountability.
4.1.3. BoundState
Definition: Marks whether AI evolution is bound to human cognitive
evolution, serving as a "switch" for symbiotic evolution. This
field integrates the responsibility closure mechanism of HJS/other
compliant accountability systems and the human judgment
verification requirement of JEP/other compliant judgment systems,
ensuring that only AI evolution bound to human cognitive evolution
can take effect, preventing unilateral evolution.
Values:
True: AI evolution is bound to human cognitive evolution
(synchronous co-evolution), and the evolution event is allowed
to take effect. The binding relationship must be verified by
JEP/other compliant judgment systems and recorded in HJS/other
compliant accountability systems to ensure accountability.
False: AI evolution is not bound to human cognitive evolution
(unilateral evolution), and the evolution event is rejected, in
line with the human participation requirement of JEP/other
compliant judgment systems and the responsibility management
principle of HJS/other compliant accountability systems.
Optional Note: A short text description (≤ 100 characters) explaining
the reason for BoundState (e.g., "Bound to JEP judgment ID: 12345,
ensuring value alignment" or "Bound to [Other Compliant System]
judgment ID: 67890, ensuring traceability"). This field is
optional but recommended for accountability and traceability,
consistent with the traceability requirement of HJS/other
compliant accountability systems and the record completeness
principle of JEP/other compliant judgment systems.
Constraints: Must be a Boolean value; the evolution event is rejected
if BoundState is False. If BoundState is True, the associated
HumanAnchor must be valid (verified by JEP/other compliant
judgment systems) and the responsibility information must be
recorded in HJS/other compliant accountability systems, ensuring
the closed-loop of binding and accountability.
4.2. Optional Extended Fields
To adapt to specific scenarios (e.g., compliance, traceability), CEP
supports optional extended fields. These fields do not change the
core structure of CEP and are compatible with the mandatory fields,
while supporting adaptation to JEP, HJS, and other compliant systems.
The design of extended fields fully considers the complementary
advantages of all compliant systems, further enhancing the
traceability and accountability of CEP.
EchoLogID (String): Unique identifier of the evolution log,
associated with the accountability log ID of HJS or other
compliant accountability systems, which can be directly
synchronized to their AI agent judgment accountability traceability
system.
EvolveTime (Timestamp): The time when the AI evolution event occurs
(format: ISO 8601), consistent with the timestamp format of JEP/
other compliant judgment systems.
OperatorID (String): Unique identifier of the human operator,
associated with the responsible person ID of HJS/other compliant
accountability systems and the judgment executor ID of JEP/other
compliant judgment systems.
SystemAdaptor (String, Optional): Unique identifier of the compliant
judgment/accountability system adapted by CEP (e.g., "JEP+HJS",
"[Other Judgment System]+[Other Accountability System]").
4.3. Protocol Frame Format
CEP uses a lightweight frame format, compatible with JSON and
Protocol Buffers (Protobuf) to adapt to different network
environments and AI systems. The core frame format inherits the
lightweight JSON/Protobuf compatibility of JEP/other compliant
judgment systems and the structured data organization advantage of
HJS/other compliant accountability systems, ensuring readability,
high performance, and compatibility with all compliant systems. The
core frame format is as follows (JSON example, including adaptation
to JEP+HJS and other systems):
{
"CEP": {
"version": "0.1",
"systemAdaptor": "JEP+HJS",
"coreFields": {
"HumanAnchor": "f47ac10b-58cc-4372-a567-0e02b2c3d479",
"ShiftClass": "EMERGENCE",
"BoundState": {
"value": true,
"note": "Bound to JEP nonce: f47ac10b-58cc-4372-a567-0e02b2c3d479"
}
},
"extendedFields": {
"EchoLogID": "log:789e4567-e89b-12d3-a456-426614174000",
"EvolveTime": "2026-03-21T10:00:00Z",
"OperatorID": "did:example:operator-98765"
}
}
}
Note: For JEP association, the HumanAnchor field SHOULD use the JEP
event's nonce (UUIDv4) as the primary identifier, as defined in
Section 2.2 of [I-D.wang-jep-judgment-event-protocol]. The multihash
"what" field MAY be used as an alternative reference when appropriate.
4.4. Protocol Interaction Process
CEP's interaction process is simple and lightweight, involving three
core roles: AI System, compliant judgment system (JEP or other), and
compliant accountability system (HJS or other). The process fully
integrates the verification advantage of compliant judgment systems
and the accountability traceability advantage of compliant
accountability systems, ensuring the validity of co-evolution binding
and the completeness of accountability closure. The process is as
follows:
1. The AI System detects an evolution event (e.g., capability
emergence, parameter fine-tuning, logic drift), determines the
compliant judgment/accountability system to be used (defaults to
JEP+HJS), generates CEP frame with mandatory core fields
(HumanAnchor, ShiftClass, BoundState) and optional extended
fields (if needed), and associates the HumanAnchor with the valid
judgment ID (e.g., JEP nonce) from the selected compliant
judgment system.
2. The AI System sends the generated CEP frame to the selected
compliant judgment system (JEP or other) to verify the validity
of the HumanAnchor field. The judgment system performs
verification in accordance with its standardized signature
verification rules (for JEP, as defined in Section 2.4 of
[I-D.wang-jep-judgment-event-protocol], including nonce
uniqueness, timestamp validity, and signature correctness) to
confirm that the associated judgment event is valid.
3. If the verification fails (e.g., invalid HumanAnchor,
unassociated judgment event), the compliant judgment system
returns a verification failure message to the AI System, and the
AI System rejects the AI evolution event, terminating the
interaction process.
4. If the verification succeeds, the compliant judgment system
returns a verification success message, and the AI System
executes the AI evolution event (only when BoundState is True; if
BoundState is False, the evolution event is directly rejected
regardless of verification result).
5. After the evolution event is executed (or rejected), the AI
System synchronizes the CEP frame (including evolution result) to
the selected compliant accountability system (HJS or other). The
accountability system records the CEP evolution information,
associates it with the corresponding judgment record and human
responsibility information from the compliant judgment system,
and completes the accountability traceability entry in accordance
with its structured responsibility management framework and
privacy architecture.
6. The compliant accountability system returns a synchronization
success message to the AI System, and the CEP interaction process
is completed.
4.5. Adaptation Specifications for Other Compliant Systems
To ensure the compatibility and reliability of CEP when adapting to
other compliant judgment/accountability systems (excluding JEP and
HJS), the following adaptation specifications must be met to avoid
conflicts with CEP's core mechanism and ensure the achievement of
human-AI symbiosis goals:
1. Core Requirement Compliance: The adapted judgment/accountability
system must meet the core requirements of human-AI symbiosis,
including but not limited to: ensuring human sovereignty in AI
evolution, providing standardized human judgment records (for
judgment systems) or structured accountability traceability (for
accountability systems), supporting non-repudiation and tamper-
proofing of data, and complying with international compliance
standards (e.g., GDPR, EU AI Act).
2. Format Compatibility: The judgment ID format of the compliant
judgment system must be compatible with CEP's HumanAnchor field
(supporting UUIDv4, multihash, or other internationally
standardized formats); the accountability log ID format of the
compliant accountability system must be compatible with CEP's
EchoLogID field, ensuring seamless association and data
synchronization.
3. Verification Mechanism Compatibility: The compliant judgment
system must provide a standardized signature verification
mechanism (including nonce uniqueness, timestamp validity,
signature correctness) consistent with CEP's verification
requirements, ensuring that the validity of the HumanAnchor field
can be verified without modifying CEP's core logic.
4. Data Synchronization Compatibility: The compliant accountability
system must support the synchronization of CEP evolution records,
be able to associate with the judgment records of the selected
compliant judgment system, and complete the entry of
accountability traceability, ensuring the closed-loop of
"judgment-binding-accountability".
5. Non-Intrusive Adaptation: The adaptation process of other
compliant systems must not modify CEP's core structure (3
mandatory core fields) and core design principles, and must be
implemented in a middleware-style deployment consistent with CEP,
avoiding intrusion into CEP's internal logic or AI model
structure.
5. Security Considerations
The security of CEP relies on the security of the underlying
compliant judgment and accountability systems (JEP, HJS, or other
adapted systems). Implementations MUST ensure:
o The integrity and authenticity of the HumanAnchor field are
verified through the signature verification mechanisms of the
associated compliant judgment system (for JEP, as defined in
Section 2.4 of [I-D.wang-jep-judgment-event-protocol]).
o The BoundState field is enforced as a critical control point; AI
evolution events with BoundState set to False MUST be rejected
unconditionally.
o All CEP frames SHOULD be transmitted over secure channels (e.g.,
TLS 1.3 or equivalent) to prevent man-in-the-middle attacks.
o Replay protection is provided by the underlying compliant judgment
system (e.g., nonce uniqueness in JEP, as defined in Section 2.3
of [I-D.wang-jep-judgment-event-protocol]) and MUST NOT be
bypassed.
6. Privacy Considerations
CEP is designed to align with privacy requirements such as GDPR,
including the right to erasure. Implementations SHOULD:
o Use the privacy architecture of HJS or other compliant
accountability systems to manage human responsibility information,
ensuring that personally identifiable information (PII) is
protected.
o Support cryptographic erasure mechanisms where required, allowing
the removal of human responsibility associations when legally
mandated.
o Minimize the collection of human-identifiable data; the
OperatorID field SHOULD be a pseudonymous identifier (e.g., DID)
rather than directly identifying information.
7. IANA Considerations
This document has no IANA actions.
8. References
8.1. Normative References
[RFC2119] Bradner, S., "Key words for use in RFCs to Indicate
Requirement Levels", BCP 14, RFC 2119,
DOI 10.17487/RFC2119, March 1997,
<https://www.rfc-editor.org/info/rfc2119>.
[RFC8174] Leiba, B., "Ambiguity of Uppercase vs Lowercase in RFC
2119 Key Words", BCP 14, RFC 8174, DOI 10.17487/RFC8174,
May 2017, <https://www.rfc-editor.org/info/rfc8174>.
[RFC8259] Bray, T., Ed., "The JavaScript Object Notation (JSON) Data
Interchange Format", STD 90, RFC 8259,
DOI 10.17487/RFC8259, December 2017,
<https://www.rfc-editor.org/info/rfc8259>.
[RFC8785] Rundgren, A., Jordan, B., and S. Erdtman, "JSON
Canonicalization Scheme (JCS)", RFC 8785,
DOI 10.17487/RFC8785, June 2020,
<https://www.rfc-editor.org/info/rfc8785>.
[RFC9562] Davis, K., Peabody, B., and P. Leach, "Universally Unique
IDentifiers (UUIDs)", RFC 9562, DOI 10.17487/RFC9562,
May 2024, <https://www.rfc-editor.org/info/rfc9562>.
[ISO8601] International Organization for Standardization, "Data
elements and interchange formats -- Information
interchange -- Representation of dates and times",
ISO 8601-1:2019, February 2019.
8.2. Informative References
[I-D.wang-jep-judgment-event-protocol]
Wang, Y., "Judgment Event Protocol (JEP)", Work in
Progress, Internet-Draft, draft-wang-jep-judgment-event-
protocol-01, 20 March 2026,
<https://datatracker.ietf.org/doc/html/draft-wang-jep-
judgment-event-protocol-01>.
[I-D.wang-hjs-judgment-structure]
Wang, Y., "Human Judgment Structure (HJS)", Work in
Progress, Internet-Draft, draft-wang-hjs-judgment-
structure-01, 20 March 2026,
<https://datatracker.ietf.org/doc/html/draft-wang-hjs-
judgment-structure-01>.
Author's Address
Yuqiang Wang
HUMAN JUDGMENT SYSTEMS FOUNDATION LTD.
Email: signal@humanjudgment.org
GitHub: https://github.com/hjs-spec