Standards Body · Foundational source, public edition · Released July 17, 2026

Canonical record: https://standardsbody.ai/library/foundational-source/project-identity/

Standards Body is an independent research and institutional-design project. It is not currently a regulator, accreditation body, certification body, or governmental authority. This document is research; it is not an adopted standard.

PROJECT_IDENTITY.md

Standards Body Project Identity

Project name: Standards Body
Primary domain: standardsbody.ai
Secondary domain: thestandardsbody.ai
Core line: Foundations for Frontier AI
Document type: Canonical project identity and positioning source
Version: 1.0
Status: Approved foundational source
Document owner: Standards Body
Applies to: Website, publications, partnerships, contributor materials, research outputs, public statements, media descriptions, institutional documents, and future organizational development
Review cycle: Annual review, with event-triggered revision after a material change in mission, institutional status, governance, authority, organizational structure, or public role

Document Purpose

This document defines the identity of Standards Body.

It is the authoritative source for:

What Standards Body is
What Standards Body is not
Why the project exists
What it is building
Which audiences it serves
How it should describe itself publicly
Which claims it may and may not make
How the present research project differs from a potential future institution
How its foundations, standards work, evaluation work, governance work, and public materials fit together
Which principles should remain stable as the project develops
How changes to project identity should be reviewed and recorded

All public descriptions of Standards Body should remain consistent with this document.

Where another project file conflicts with this document on matters of identity, present authority, mission, public positioning, or institutional status, this document governs unless it has been formally superseded.

1. Canonical Identity Statement

1.1 Primary Identity

Standards Body is an independent research and institutional-design project developing foundations for frontier AI evaluation, assurance, standards, and governance.

1.2 Extended Identity

Standards Body develops the concepts, methods, institutional frameworks, and shared language needed for credible frontier AI evaluation and standards. Its work focuses on dynamic evaluation, held-out testing, high-stakes capability assessment, independent review, third-party assurance, progressive requirements, incentive alignment, and global interoperability.

1.3 Present Status

Standards Body is presently:

An independent project
A research initiative
An institutional-design effort
A developing knowledge base
A source of conceptual frameworks
A future-facing standards initiative
A platform for structured inquiry, documentation, and contribution

Standards Body is not presently:

A government agency
A regulator
A statutory standards authority
An accreditation body
A certification body
A conformity-assessment body
An official international organization
A licensing authority
A legal enforcement body
A representative of any government
A representative of the artificial intelligence industry
An organization authorized to approve or prohibit AI systems
A source of legal determinations

1.4 Core Line

Foundations for Frontier AI

This line expresses the project's role.

Standards Body is building the intellectual and institutional foundations required before credible frontier AI standards can mature.

It does not imply that the project has already established binding standards.

1.5 Central Identity Boundary

Standards Body may research, propose, organize, convene, evaluate, publish, and develop frameworks. It may not imply that its work has legal, regulatory, accreditation, certification, or governmental authority unless such authority is later lawfully and explicitly established.

2. Why Standards Body Exists

2.1 The Institutional Gap

Frontier artificial intelligence is developing faster than the institutions responsible for evaluating and governing it.

Current systems often depend on fragmented arrangements:

Developers evaluate their own models
Public benchmarks become optimized and stale
External evaluators receive inconsistent access
Capability results are difficult to compare
Evaluation methods change without common version control
High-stakes claims rely on incomplete evidence
Auditing, certification, accreditation, and review are used imprecisely
Voluntary practices and legal requirements are discussed as though they are the same
International institutions use incompatible terminology
Incident evidence does not consistently feed back into standards
Public claims exceed what evaluation results can support
Small and independent evaluators lack infrastructure and recognition
Standards processes can move too slowly for rapidly changing systems
Fast-moving technical work can lack institutional legitimacy
Formal governance can harden before the science is ready

Standards Body exists to help close this gap.

2.2 The Foundational Problem

A credible standards system cannot begin with a badge, certification mark, or regulatory claim.

It must begin with:

Definitions
first principles
evidence standards
evaluation philosophy
research methods
institutional roles
governance
conflict controls
transparency
correction
participation
versioning
interoperability

The project begins at this foundational layer.

2.3 The Long-Term Need

Frontier AI will require institutions capable of answering questions such as:

What should be evaluated?
Which methods are valid?
How should evidence be protected?
Who is competent to evaluate?
When is external review required?
How should conflicts be managed?
What does a threshold mean?
When should voluntary practice become a formal requirement?
How can evaluators be recognized across jurisdictions?
How should incidents update standards?
How should standards remain current?
How can public trust be earned without overclaiming?

Standards Body is developing a coherent architecture for these questions.

2.4 The Project's Response

Standards Body responds by building:

A first-principles foundation
A common intellectual framework
A standards-development architecture
An evaluator and assurance model
A transparent institutional design
A living knowledge base
A pathway toward future legitimate standards work

3. Mission

3.1 Canonical Mission

The mission of Standards Body is to develop credible foundations for frontier AI evaluation, assurance, standards, and governance in the public interest.

3.2 Mission in Practice

Standards Body advances this mission by:

Defining clear concepts and terminology
Developing evaluation principles and protocols
Studying high-stakes capabilities and safeguards
Strengthening independent review
Designing accountable third-party assurance systems
Exploring how standards should mature over time
Aligning incentives with evidence and integrity
Supporting international interoperability
Publishing open questions and research priorities
Building transparent processes for contribution, correction, and revision
Distinguishing technical evidence from legal and political authority

3.3 Mission Constraints

The mission does not include:

Declaring all advanced AI dangerous
Declaring frontier systems safe
Acting as a substitute for democratic government
Creating private regulatory power without accountability
Promoting one company, model, ideology, or jurisdiction
Treating compliance as proof of safety
Treating technical expertise as a complete source of political legitimacy
Freezing immature evaluation science into permanent rules
Using standards language as institutional theater

3.4 Public-Interest Orientation

Standards Body's work should support:

Better evidence
safer beneficial development
accountable institutions
credible evaluation
informed public decisions
international cooperation
plural participation
correction of error
reduction of unnecessary duplication
protection against institutional capture

The project does not define the public interest as opposition to technological progress.

It defines the public interest as the responsible development and use of frontier AI under conditions that make important claims reviewable, evidence-based, and accountable.

4. Vision

4.1 Long-Term Vision

A world in which consequential claims about frontier AI are supported by credible, current, independently reviewable, and internationally interpretable evidence.

4.2 Institutional Vision

The long-term institutional vision includes an ecosystem in which:

Evaluation protocols remain current
Protected tests preserve evidentiary value
High-stakes capability claims receive proportionate scrutiny
Independent reviewers have adequate access
Evaluator organizations demonstrate competence and impartiality
Standards mature through transparent stages
Incentives reward integrity and public contribution
Evidence can travel across institutions and borders
Incidents improve future evaluation
Standards can be corrected and retired
Public claims remain within evidence
Authority is clearly bounded
Technical and political decisions remain appropriately separated

4.3 Vision Without Presumption

Standards Body does not presume that it must become the sole institution performing these functions.

Its work should remain useful whether future responsibilities are carried by:

Standards organizations
Governments
National AI institutes
International organizations
Independent nonprofits
Academic consortia
Accredited evaluators
Industry bodies
Public-interest institutions
New institutional forms

4.4 Preferred Future

The preferred future is plural and interoperable.

No single developer, evaluator, government, or standards organization should become the unquestionable authority over all frontier AI evidence.

5. The Problem Standards Body Is Solving

5.1 The Evaluation Problem

Frontier AI evaluation faces:

Rapid capability change
benchmark contamination
weak real-world validity
inconsistent elicitation
scarce domain expertise
model and system complexity
uncertain thresholds
strategic or adaptive behavior
incomplete deployment context
insufficient post-deployment evidence

5.2 The Assurance Problem

Independent assurance faces:

Developer-controlled access
inconsistent reviewer competence
financial dependence
unclear scopes
inadequate security
weak public reporting
evaluator shopping
absence of proficiency testing
unclear accreditation pathways

5.3 The Standards Problem

Standards development faces:

Immature measurement science
rapidly changing objects
slow consensus processes
incumbent influence
international fragmentation
poor retirement mechanisms
compliance theater
confusion between voluntary and mandatory requirements

5.4 The Governance Problem

Governance faces:

Concentrated information
concentrated decision power
unclear accountability
weak correction
insufficient dissent
competing national interests
underrepresentation of smaller institutions and regions
pressure for premature certainty

5.5 The Communication Problem

Public understanding is weakened when:

Evaluation results are described as proof of safety
Reviews are described as audits without defined criteria
Advisory findings are described as certifications
Private frameworks are described as regulation
Model names conceal material configuration differences
Uncertainty is omitted
Institution names imply authority not actually held

Standards Body treats precise institutional language as part of substantive integrity.

6. What Standards Body Builds

6.1 Foundational Research

Standards Body develops first-principles papers concerning:

Dynamic evaluation protocols
Held-out evaluations
High-stakes capability evaluation
Independent expert review
Third-party auditor ecosystems
Progressive standards and requirements
Incentives and prestige
Global interoperability

6.2 Intellectual Infrastructure

The project develops:

Terminology
taxonomies
evidence standards
research methodology
evaluation philosophy
open questions
research agendas
knowledge-gap records
contrarian arguments
source registries

6.3 Institutional Infrastructure

The project develops:

Institution design
governance frameworks
standards-development processes
evaluator accreditation concepts
contributor systems
transparency rules
partnership principles
long-term roadmaps

6.4 Operational Templates

Standards Body may publish templates for:

Evaluation protocols
evidence cases
reviewer mandates
conflict disclosures
evaluator profiles
accreditation scopes
incident records
standards proposals
crosswalks
recognition decisions
version histories
public reports

6.5 Living Knowledge Base

The project maintains:

Sources
timelines
person and institution profiles
case studies
failure databases
early signals
future work
version histories
website source of truth

6.6 Public Explanation

Standards Body translates technical and institutional questions into:

Clear articles
explainers
visual frameworks
public summaries
briefing materials
structured comparisons

6.7 Future Pilots

Subject to competence, resources, security, and governance, Standards Body may support limited pilots involving:

Protocol design
cross-institution comparison
evaluator qualification
standards-development exercises
reporting frameworks
evidence interoperability
contributor recognition

A pilot does not itself establish formal authority.

7. What Standards Body Does Not Build

Standards Body should not present itself as building:

A universal AI safety score
A single global regulator
A system for approving all AI models
A blacklist of companies or technologies
A permanent central authority
A private lawmaking body
A public-relations certification program
An ideological campaign against frontier AI
A promotional vehicle for frontier laboratories
A substitute for sector-specific expertise
A substitute for national law
A substitute for public participation
A claim that every advanced capability is high risk
A claim that every risk can be measured
A claim that standards eliminate uncertainty

8. Core Principles

8.1 Evidence Before Authority

Institutional claims should not exceed the evidence supporting them.

8.2 Independence Without Isolation

Independent scrutiny should combine genuine judgment with sufficient system access and developer cooperation.

8.3 Proportionality

Evaluation and assurance should become more rigorous as consequence and uncertainty increase.

8.4 Revisability

Protocols, standards, and institutional positions should be versioned, correctable, and retireable.

8.5 Transparency With Security

Public accountability and protected information must be balanced through explicit disclosure levels.

8.6 Precision of Language

Terms such as standard, audit, certification, accreditation, recognition, and regulation should be used only when their institutional meaning is justified.

8.7 Public Interest

The work should improve the quality of decisions affecting people, institutions, and society.

8.8 Pluralism

No single technical discipline, institution, geography, ideology, or stakeholder group should control the complete agenda.

8.9 Global Interoperability

Shared evidence should support cooperation without erasing legitimate local authority and difference.

8.10 Accountability

Developers, evaluators, standards bodies, and Standards Body itself should be subject to challenge, correction, and documented decision processes.

8.11 Anti-Capture

The project should actively examine financial, institutional, ideological, professional, and geopolitical conflicts.

8.12 Contribution Over Prestige

Recognition should follow demonstrated contribution rather than status alone.

8.13 Safety and Benefit

The project should support beneficial frontier AI development while strengthening the evidence and safeguards appropriate to consequential capabilities.

8.14 Institutional Humility

Standards Body should state what it does not know and avoid performing authority it has not earned.

8.15 Show the Work

Important conclusions should be traceable to:

Sources
assumptions
methods
evidence
uncertainty
review
version

9. Foundational Positioning

9.1 The Project Is Pro-Evidence

Standards Body is not defined by optimism or pessimism about AI.

It is defined by the position that consequential claims require credible evidence.

9.2 The Project Is Pro-Institution

Standards Body recognizes that technical systems require institutional infrastructure.

It does not assume that markets, laboratories, governments, or civil society can solve every problem alone.

9.3 The Project Is Pro-Progress

The project supports beneficial scientific and technological progress.

It rejects the assumption that supporting progress requires weak evaluation or that supporting safeguards requires opposition to innovation.

9.4 The Project Is Non-Aligned

Standards Body should not be institutionally aligned with:

A specific frontier laboratory
A specific political party
A national industrial policy
A single school of AI-risk thought
A proprietary evaluator
A certification business
An anti-technology movement

This does not require neutrality toward evidence.

9.5 The Project Is Non-Alarmist

Standards Body should discuss severe possibilities proportionately and precisely.

It should avoid:

Sensational predictions
unsupported certainty
fear-based branding
treating speculative scenarios as established facts
equating all frontier progress with catastrophe

9.6 The Project Is Not Complacent

Non-alarmism does not mean minimizing credible high-consequence risk.

Where the consequences of error are large, the project supports stronger evidence, review, safeguards, and institutional preparation.

9.7 The Project Is Institutionally Serious

Public materials should communicate:

Calm authority
careful distinction
durable thinking
high standards of evidence
willingness to revise
respect for complexity

10. Present Institutional Stage

10.1 Current Stage Name

Foundational research and institutional design

10.2 Current Activities

At this stage, Standards Body may:

Publish research
Develop frameworks
Build taxonomies
Draft standards concepts
Create templates
Maintain source registries
Invite expert critique
Convene exploratory discussions
Develop pilot proposals
Build public educational materials
Establish project governance

10.3 Current Non-Activities

At this stage, Standards Body should not:

Certify AI systems
Accredit evaluators
Issue official standards-conformity findings
Grant regulatory approval
Claim international recognition
Issue legal compliance opinions
Publish an official safety rating
Impose binding obligations
Use seals or marks implying formal assurance

10.4 Stage Transition

Any move into a materially different institutional role should require:

A written institutional proposal
competence analysis
governance review
conflict analysis
resource assessment
public-interest justification
authority review
security assessment
transparency plan
versioned amendment to this document

10.5 No Implied Future Authority

References to a possible future standards institution are exploratory.

They do not create present authority or guarantee that Standards Body will adopt that role.

11. Future Institutional Possibilities

Standards Body may study or prepare for future roles such as:

Research institute
Standards-development organization
Evaluation-methods steward
Independent nonprofit
Public-interest technical secretariat
Convening body
Protocol registry
Knowledge infrastructure provider
International interoperability forum
Evaluator-support institution

More formal possibilities may be examined only after adequate governance and competence exist.

These include:

Scheme ownership
evaluator recognition
formal standards publication
certification-related support
accreditation partnership
controlled evidence exchange

No such role should be publicly claimed before formal establishment.

12. Audiences

12.1 Frontier AI Developers

Standards Body provides:

Evaluation architecture
review principles
standards pathways
assurance concepts
interoperability frameworks
external challenge

The project should neither treat developers as enemies nor permit developer access to determine conclusions.

12.2 Independent Evaluators and Auditors

Standards Body provides:

Competence concepts
scope models
quality principles
conflict controls
recognition pathways
common terminology
operational templates

12.3 Researchers

Standards Body provides:

Research agendas
evidence standards
open questions
benchmark and protocol concepts
institutional context
public-goods recognition

12.4 Governments and Public Institutions

Standards Body provides:

Technical frameworks
institutional analysis
standards pathways
evidence structures
interoperability concepts

It does not present its work as official policy.

12.5 Standards and Accreditation Organizations

Standards Body provides:

Frontier-specific research
dynamic methods
terminology
institutional crosswalks
evidence about emerging assurance needs

12.6 Purchasers, Insurers, and Investors

Standards Body provides:

Clear distinctions among assurance claims
evaluator-quality concepts
evidence requirements
procurement and risk frameworks

It does not provide investment, insurance, or legal advice.

12.7 Open-Source and Open-Weight Communities

Standards Body provides:

Portable evaluation concepts
contribution pathways
model-identity frameworks
recognition of decentralized expertise
institutional representation principles

12.8 Civil Society and Public-Interest Organizations

Standards Body provides:

Public accountability frameworks
transparency principles
institutional analysis
participation pathways
evidence interpretation

12.9 Journalists and Public Communicators

Standards Body provides:

Precise terminology
bounded claims
public explainers
institutional distinctions
source-based analysis

12.10 The General Public

Standards Body should make its core ideas understandable without reducing complexity to slogans or fear.

13. Primary Areas of Work

13.1 Evaluation Science

Questions include:

What should be measured?
How should capability be elicited?
How should uncertainty be reported?
How should protocols change?
What evidence supports a threshold?

13.2 Independent Review

Questions include:

What makes review independent?
What access is necessary?
How should dissent be preserved?
How should conflicts be managed?

13.3 Assurance Ecosystems

Questions include:

Who evaluates evaluators?
How should competence be scoped?
When is certification meaningful?
How should accreditation adapt?

13.4 Standards Development

Questions include:

When is a practice mature enough for a standard?
How should standards be tested?
How should they be revised and retired?
How should capture be prevented?

13.5 Institutional Governance

Questions include:

Who decides?
Who reviews?
Who may appeal?
How should authority remain bounded?
How should emergency decisions work?

13.6 Incentive Alignment

Questions include:

Which work is rewarded?
How should correction receive credit?
How should public goods be funded?
How can prestige track contribution?

13.7 Global Interoperability

Questions include:

How should evidence travel?
How should protocols be mapped?
How should evaluator competence be recognized?
How should local authority be preserved?

13.8 Public Knowledge Infrastructure

Questions include:

Which sources are current?
Which terms conflict?
Which incidents reveal failure?
Which early signals matter?
Which open questions remain?

14. The Eight Foundational Papers

The initial foundational library consists of:

FOUNDATION_01_DYNAMIC_EVALUATION_PROTOCOLS.md
FOUNDATION_02_HELD_OUT_EVALUATIONS.md
FOUNDATION_03_HIGH_STAKES_CAPABILITY_EVALUATION.md
FOUNDATION_04_INDEPENDENT_EXPERT_REVIEW.md
FOUNDATION_05_THIRD_PARTY_AUDITOR_ECOSYSTEM.md
FOUNDATION_06_PROGRESSIVE_STANDARDS_AND_REQUIREMENTS.md
FOUNDATION_07_INCENTIVES_AND_PRESTIGE.md
FOUNDATION_08_GLOBAL_INTEROPERABILITY.md

Together they define the project's initial view of a credible frontier AI standards architecture.

14.1 Foundation Relationship

The foundations form a sequence:

Protocols must remain current.
Some evaluation evidence must remain protected.
High-stakes capabilities require proportionate scrutiny.
Consequential claims require independent review.
Independent review must scale through competent institutions.
Practices should mature into stronger requirements only when justified.
Incentives must reward integrity and contribution.
Evidence must remain usable across institutions and borders.

14.2 Foundation Status

The foundation papers are:

Canonical working white papers
Versioned
Revisable
Subject to expert challenge
Not legal standards
Not regulatory instruments
Not certifications
Not declarations of universal consensus

15. Relationship to Existing Institutions

15.1 Governments

Standards Body may support public institutions with research and frameworks.

It does not speak for governments.

15.2 Regulators

Standards Body may analyze regulatory questions.

It does not make binding regulatory decisions.

15.3 Standards Organizations

Standards Body may contribute research and proposals.

It should not claim equivalence with formally recognized standards organizations unless such status is established.

15.4 Accreditation Bodies

Standards Body may study accreditation.

It does not accredit.

15.5 Certification Bodies

Standards Body may study certification schemes.

It does not certify.

15.6 AI Developers

Standards Body may collaborate with developers for access, evidence, and technical understanding.

Collaboration should not purchase conclusions or control identity.

15.7 Evaluators

Standards Body may support evaluator methods and institutional development.

It should not describe an evaluator as formally approved unless a legitimate process exists.

15.8 Academic Institutions

Standards Body may collaborate on research.

Academic affiliation does not automatically validate project claims.

15.9 Civil Society

Standards Body should include public-interest perspectives and affected-party knowledge.

15.10 International Organizations

Standards Body may participate in or support international discussions.

It should not imply membership, recognition, or endorsement without evidence.

16. Independence

16.1 Independence Position

Standards Body should remain sufficiently independent to:

Critique developers
critique regulators
critique evaluators
critique standards bodies
critique its own work
publish uncertainty
preserve dissent
correct conclusions

16.2 Independence Dimensions

The project should monitor:

Organizational independence
financial independence
methodological independence
publication independence
intellectual independence
political independence
security independence

16.3 Funding

Funding sources should be disclosed according to the future transparency framework.

No funder should receive:

Control over conclusions
veto over publication
hidden influence over standards
automatic governance authority
favorable treatment
recognition beyond contribution

16.4 Partnerships

Partnerships should be:

Purpose-specific
documented
conflict-reviewed
bounded
publicly described where appropriate
terminable

16.5 Independence Is Not Non-Engagement

Standards Body may engage deeply with institutions it evaluates or studies.

The requirement is not distance.

The requirement is freedom of judgment and transparent conflict management.

17. Political and Ideological Position

17.1 Nonpartisan Position

Standards Body should remain nonpartisan.

Its work may have policy implications.

That does not make it a political-party organization.

17.2 Evidence-Based Disagreement

The project may support or oppose a proposal when evidence and institutional reasoning justify a position.

Nonpartisanship does not require silence.

17.3 Ideological Pluralism

Standards Body should not define participation by adherence to:

Accelerationism
effective altruism
technological pessimism
technological optimism
open-source ideology
closed-model ideology
national industrial strategy
anti-regulatory ideology
pro-regulatory ideology

Contributors should be evaluated by relevance, evidence, competence, integrity, and conduct.

17.4 Values Are Explicit

The project is not value-free.

It values:

Evidence
accountability
public benefit
human agency
institutional legitimacy
correction
pluralism
cooperation
beneficial progress

18. Voice and Tone

18.1 Canonical Voice

Standards Body should sound:

Calm
authoritative
precise
serious
accessible
independent
institutionally mature
non-alarmist
evidence-aware
willing to state uncertainty

18.2 Avoided Voice

Standards Body should not sound:

Sensational
promotional
combative
apocalyptic
self-important
bureaucratic without substance
falsely neutral
politically tribal
overly academic
startup-hyped
technologically naive

18.3 Public Writing Principles

Public writing should:

Define important terms
separate fact from inference
state uncertainty
identify institutional limits
avoid unsupported superlatives
avoid claims of consensus without evidence
avoid implying official authority
explain why distinctions matter
use direct sentences
remain readable without losing precision

18.4 First-Person Use

Use "Standards Body" for formal institutional positions.

Use "we" sparingly in public-facing writing where a human and accountable institutional voice is appropriate.

18.5 Tone Under Disagreement

Disagreement should be expressed through:

Evidence
reasoning
institutional analysis
clear boundaries

Avoid personal attacks and motive speculation without evidence.

19. Naming and Brand Architecture

19.1 Official Project Name

Standards Body

19.2 Approved Extended Form

Standards Body, Foundations for Frontier AI

19.3 Primary Domain

standardsbody.ai

19.4 Secondary Domain

thestandardsbody.ai

19.5 Core Line

Foundations for Frontier AI

19.6 Name Interpretation

The name expresses the long-term subject and institutional ambition of the project.

It does not claim present statutory status.

19.7 Name Usage Rule

Public materials should include a clear present-status description where the name could reasonably be interpreted as an official regulator or formally recognized standards authority.

19.8 Prohibited Name Extensions

Do not use without formal authorization:

International Standards Body
Global AI Standards Authority
Official AI Standards Body
Frontier AI Regulator
AI Accreditation Authority
AI Certification Authority
Government Standards Body
United Nations Standards Body
Federal Standards Body

19.9 Product and Program Names

Programs should use clear descriptive names and avoid authority inflation.

Preferred:

Standards Body Research Program
Standards Body Foundation Series
Standards Body Evaluation Protocol Pilot
Standards Body Contributor Program
Standards Body Interoperability Project

Avoid:

Official Approval Program
Global Certification Authority
Regulatory Seal
Licensed AI Program

20. Approved Public Descriptions

20.1 One-Sentence Description

Standards Body is an independent research and institutional-design project developing foundations for frontier AI evaluation, assurance, standards, and governance.

20.2 Two-Sentence Description

Standards Body is an independent project developing the concepts, methods, and institutional frameworks needed for credible frontier AI evaluation and standards. Its work focuses on dynamic protocols, held-out testing, high-stakes capabilities, independent review, third-party assurance, progressive requirements, incentive alignment, and global interoperability.

20.3 Short Website Description

Standards Body develops foundations for credible frontier AI evaluation, assurance, standards, and governance. We are building the shared language, evidence principles, institutional frameworks, and public knowledge needed before durable standards can mature.

20.4 Extended Website Description

Standards Body is an independent research and institutional-design project focused on frontier artificial intelligence. We develop foundational work on evaluation protocols, protected testing, high-stakes capability assessment, independent expert review, third-party assurance, progressive standards, incentive alignment, and global interoperability. Standards Body is not a regulator, accreditation body, or certification authority. Our current role is to build the intellectual and institutional foundations required for more credible standards work.

20.5 Publication Boilerplate

About Standards Body: Standards Body is an independent research and institutional-design project developing foundations for frontier AI evaluation, assurance, standards, and governance. Its work is research and framework development and does not constitute certification, accreditation, regulatory approval, or legal advice.

20.6 Partnership Boilerplate

Standards Body is an independent frontier AI standards research project. Participation or collaboration does not imply endorsement, certification, accreditation, regulatory recognition, or agreement with every Standards Body position.

20.7 Media Description

Standards Body is a developing independent initiative focused on the infrastructure behind credible frontier AI standards, including evaluation methods, independent review, assurance systems, and international interoperability.

20.8 Social Profile Description

Independent research and institutional design for frontier AI evaluation, assurance, and standards. Foundations for Frontier AI.

21. Approved Authority Disclaimers

21.1 General Disclaimer

Standards Body is an independent research and institutional-design project. It is not a government agency, regulator, accreditation body, certification body, or legal authority.

21.2 Publication Disclaimer

This publication presents research, analysis, and proposed frameworks. It does not constitute a legal standard, certification decision, accreditation decision, regulatory determination, or guarantee of system safety.

21.3 Evaluation Disclaimer

An evaluation result describes evidence under specified conditions. It should not be interpreted as proof of universal capability, incapability, safety, legal compliance, or future behavior.

21.4 Partnership Disclaimer

Collaboration with Standards Body does not imply institutional endorsement, conformity, approval, certification, accreditation, or regulatory status.

21.5 Contributor Disclaimer

Contributor participation does not imply that the contributor or their organization endorses every Standards Body position.

22. Prohibited Public Claims

Standards Body should not claim:

"We regulate frontier AI."
"We are the official standards body for AI."
"We certify AI systems as safe."
"We accredit AI auditors."
"Our standards are legally binding."
"This model has been approved by Standards Body."
"This evaluator is officially licensed by Standards Body."
"Our framework guarantees safety."
"Our work represents global consensus."
"We speak for the AI industry."
"We speak for governments."
"We have determined that a company is legally compliant."
"Our evaluation proves that a capability does not exist."
"Our score determines whether a system may be deployed."
"Standards Body recognition is equivalent to government approval."

These claims may become appropriate only after a lawful and formally documented institutional change, and only within the resulting authority.

23. Claims Standards

23.1 Institutional Claims

Every institutional claim should answer:

What role is being claimed?
What authority supports it?
What scope applies?
What limitations remain?
Is the claim current?

23.2 Technical Claims

Every technical claim should identify:

System
version
protocol
conditions
evidence
uncertainty
limitations
date

23.3 Consensus Claims

Do not say "experts agree" unless:

The relevant expert community is defined
Evidence of agreement exists
Significant dissent is represented
Scope is bounded

23.4 Global Claims

Do not say "global standard," "internationally recognized," or "worldwide consensus" without documented institutional basis.

23.5 Safety Claims

Prefer:

"The system met the specified requirement under the assessed conditions."
"The evaluation did not demonstrate the defined capability under the tested conditions."
"The safeguard reduced the measured attack success rate under the stated threat model."

Avoid:

"The system is safe."
"The capability does not exist."
"The safeguard cannot be bypassed."

24. Evidence and Research Position

24.1 Evidence Hierarchy

Standards Body should prefer:

Primary technical sources
Official standards
official legal texts
direct evaluation evidence
peer-reviewed research
transparent institutional reports
documented expert judgment

24.2 Current Information

Time-sensitive institutional, technical, legal, and standards claims should be verified before publication.

24.3 Confidential Evidence

Confidential evidence may inform a conclusion only when:

Source and access are documented
reviewers are qualified
conflicts are controlled
public claims remain supportable
security is protected
limits are disclosed

24.4 Model-Generated Work

AI tools may assist research and drafting.

They should not be treated as independent evidence.

Human responsibility remains with the publication owner.

24.5 Correction

Material error should be:

Acknowledged
corrected
versioned
explained
propagated to dependent documents

24.6 Uncertainty

Uncertainty is an output, not a failure.

25. Contributor Identity

25.1 Contributor Definition

A contributor is a person or institution that provides documented work to the project.

25.2 Contributor Does Not Mean Representative

A contributor does not automatically speak for Standards Body.

25.3 Contributor Does Not Mean Endorser

Participation does not imply agreement with every position.

25.4 Contributor Recognition

Credit should reflect actual contribution.

25.5 Contributor Authority

Authority should be granted by role and governance, not prestige alone.

25.6 Conflicts

Contributors should disclose material conflicts relevant to their work.

25.7 Conduct

Contributors should engage through:

Evidence
professional conduct
respect for disagreement
confidentiality where required
correction of error

Detailed rules belong in CONTRIBUTOR_FRAMEWORK.md.

26. Partnership Identity

26.1 Partnership Purpose

Partnerships may support:

Research
evaluation
infrastructure
standards work
capacity building
translation
international cooperation
public education

26.2 Partnership Does Not Mean Endorsement

A partnership does not mean:

Agreement with every institutional position
certification
accreditation
legal recognition
regulatory approval
exclusive affiliation

26.3 Partnership Review

Before announcement, evaluate:

Purpose
conflict
independence
funding
authority
public language
security
termination

26.4 Logo and Name Use

Partners should not use the Standards Body name to imply approval beyond the written partnership scope.

26.5 Exclusive Partnerships

Avoid exclusivity where it would:

Compromise independence
restrict plural participation
create evaluator dependence
imply alignment
weaken public trust

Detailed principles belong in PARTNERSHIP_PRINCIPLES.md.

27. Funding Identity

27.1 Funding Purpose

Funding should support:

Research
infrastructure
expert contribution
public goods
security
translation
open access
institutional capacity

27.2 Funding Boundaries

Funding should not purchase:

Findings
standards outcomes
favorable treatment
publication veto
hidden influence
governance control without formal process
certification or recognition

27.3 Funding Disclosure

Material funding should be disclosed according to the transparency framework.

27.4 Funding Diversity

The project should seek funding diversity to reduce dependence.

27.5 In-Kind Support

Access to models, compute, data, facilities, or personnel is a material contribution and potential conflict.

28. Public Trust Position

28.1 Trust Is Earned

Standards Body should not ask for trust based only on:

Name
design
expert affiliation
institutional language
confidence
prestige

Trust should follow:

Evidence
transparent methods
correction
competence
independence
bounded claims
public accountability

28.2 No Institutional Theater

The project should not imitate the appearance of a mature authority before building the systems that make authority credible.

28.3 Transparency Limits

Trust does not require public release of every sensitive artifact.

It requires clarity about:

What is withheld
why it is withheld
who reviewed it
what the public claim can support

28.4 Trust and Dissent

Visible disagreement can strengthen justified trust.

29. Relationship to Regulation

29.1 Research Role

Standards Body may provide research relevant to regulation.

29.2 No Regulatory Substitution

Technical standards should not replace accountable public decision-making where legal authority is required.

29.3 Standards in Law

Standards may support:

Technical implementation
evidence
conformity
interoperability
procurement

Their incorporation into law requires public authority and due process.

29.4 Regulatory Critique

Standards Body may identify:

Measurement problems
implementation gaps
capture risks
evaluator shortages
false precision
international conflicts

29.5 Policy Neutrality Is Not Required

The project may recommend institutional action when supported by evidence.

It should distinguish technical findings from policy judgment.

30. Relationship to Industry

30.1 Constructive Engagement

Industry holds essential technical knowledge and implementation capacity.

30.2 Independent Judgment

Industry participation should not determine conclusions.

30.3 No Anti-Industry Identity

Standards Body is not organized around opposition to AI companies.

30.4 No Industry-Association Identity

Standards Body does not represent collective industry interests.

30.5 Evidence Expectations

Developers should support:

Model identity
evaluation access
internal evidence
safeguard testing
incident learning
external review
correction

30.6 Commercial Benefit

Commercial benefit from standards is not inherently improper.

It should not control standard content or public claims.

31. Relationship to Open-Source Communities

31.1 Equal Legitimacy of Evidence

Open-source contributions should be evaluated by evidence and competence.

31.2 No Ideological Presumption

Open systems are not inherently safe or unsafe.

Closed systems are not inherently safe or unsafe.

31.3 Distinct Institutional Needs

Open-weight systems require attention to:

Model lineage
modification
decentralized deployment
community governance
downstream evaluation
sustainable maintenance
distributed responsibility

31.4 Participation

Open-source experts should receive meaningful roles in relevant standards work.

31.5 Avoiding Tokenism

Participation should include:

Access
contribution authority
review
dissent
credit

32. Relationship to the Public

32.1 Public Legibility

Standards Body should make important distinctions understandable.

32.2 No Fear Marketing

The project should not build attention through panic.

32.3 No False Reassurance

The project should not minimize uncertainty to make technology appear settled.

32.4 Public Participation

Public and affected-party input is relevant to:

Deployment context
institutional legitimacy
rights
access
distributional impact
practical failure

32.5 Technical Boundaries

Public participation does not require treating every technical claim as a matter of preference.

32.6 Education

Standards Body should explain:

What evaluation can establish
What standards mean
What certification means
What uncertainty remains
Who has authority

33. Website Identity Requirements

The initial website should communicate five facts immediately:

Standards Body is focused on frontier AI.
Its present work is foundational research and institutional design.
It develops frameworks for evaluation, assurance, standards, and governance.
It is independent.
It is not presently a regulator, accreditor, or certification authority.

33.1 Homepage Priority

The homepage should prioritize:

Name
Core line
Clear identity sentence
Foundations
Mission
Present status
Public work
Contact or participation

33.2 Avoided Homepage Signals

Avoid:

Government-style seals
legalistic approval language
"certified" badges
claims of global authority
countdowns or crisis framing
vague startup language
unsupported partner logos
inflated metrics

33.3 Website Source of Truth

Exact public copy belongs in WEBSITE_SOURCE_OF_TRUTH.md.

34. Publication Identity Requirements

Every major Standards Body publication should include:

Project name
Document title
Version
Status
Date
Document purpose
Authority note
Review or revision status
Sources where applicable
Revision record

34.1 Status Labels

Approved status labels:

Exploratory note
Research brief
Working paper
Canonical working white paper
Proposed framework
Draft protocol
Pilot specification
Approved foundational source
Superseded
Withdrawn
Archived

34.2 Prohibited Status Labels Without Formal Basis

Official standard
International standard
Certified protocol
Regulatory guidance
Binding requirement
Accredited method
Government-approved framework

35. Institutional Identity Tests

Before publication or partnership, ask:

35.1 Authority Test

Could a reasonable reader believe Standards Body has official authority it does not hold?

35.2 Evidence Test

Does the claim exceed the supporting evidence?

35.3 Independence Test

Could funding, access, or partnership reasonably be seen as controlling the conclusion?

35.4 Scope Test

Is the claim limited to the actual system, method, domain, and period?

35.5 Public-Interest Test

Does the activity advance credible public value rather than status alone?

35.6 Capture Test

Does the activity give one institution disproportionate influence?

35.7 Reversibility Test

Can the decision, claim, designation, or partnership be corrected or ended?

35.8 Legibility Test

Can a non-specialist understand what Standards Body is and is not claiming?

35.9 Security Test

Does transparency create avoidable harm or exposure?

35.10 Identity Consistency Test

Is the language consistent with this document?

36. Public FAQ

36.1 Is Standards Body a regulator?

No. Standards Body is an independent research and institutional-design project.

36.2 Does Standards Body issue official AI standards?

Not at its present stage. It develops foundational research, proposed frameworks, and standards architecture.

36.3 Does Standards Body certify AI systems?

No.

36.4 Does Standards Body accredit AI evaluators?

No.

36.5 Is Standards Body affiliated with a government?

Not by default. Any specific government partnership or funding should be disclosed.

36.6 Is Standards Body affiliated with an AI company?

Not by default. Any specific partnership or funding should be disclosed and governed.

36.7 Is Standards Body opposed to frontier AI?

No. Standards Body supports beneficial progress under stronger systems of evidence, evaluation, assurance, and accountability.

36.8 Does Standards Body believe frontier AI is safe?

Standards Body does not issue broad declarations that frontier AI as a category is safe or unsafe. It evaluates claims, methods, risks, safeguards, and institutions within defined contexts.

36.9 Who can contribute?

Relevant contributors may include researchers, developers, evaluators, standards professionals, domain experts, open-source contributors, public-interest organizations, and affected communities, subject to competence, conduct, conflict, and governance requirements.

36.10 Are Standards Body papers final?

They are canonical and versioned, but revisable when evidence, systems, or institutional understanding changes.

36.11 Does Standards Body speak for the public?

No single private institution speaks for the public. Standards Body works in the public interest and should support accountable public participation and government authority where appropriate.

36.12 Why use the name Standards Body now?

The name expresses the project's subject and institutional direction. Clear public descriptions and disclaimers must distinguish the current foundational project from any future formal role.

37. Media Response Guide

Question: What is Standards Body?

Standards Body is an independent research and institutional-design project developing foundations for frontier AI evaluation, assurance, standards, and governance.

Question: Are you trying to regulate AI?

Standards Body is not a regulator. Our work focuses on the technical and institutional foundations that credible standards and public policy may eventually rely upon.

Question: Do you certify AI models?

No. We currently conduct research and framework development. We do not certify or approve AI systems.

Question: Are you affiliated with the major AI laboratories?

Standards Body is independent. Any specific funding, research access, or partnership should be disclosed and governed through conflict and transparency rules.

Question: Are you focused on catastrophic AI risk?

Our work covers high-stakes frontier capabilities as well as the institutional systems needed for credible evaluation. We approach severe risks proportionately and without sensationalism.

Question: Are you anti-innovation?

No. Credible evaluation and standards can support beneficial innovation by improving confidence, comparability, and accountability.

Question: Why is a new institution needed?

Frontier AI evaluation is fragmented, rapidly changing, and difficult to interpret across institutions. Standards Body is developing the shared foundations needed for more credible and interoperable standards work.

Plain-language summaries and claims boundaries.

39. Mission Test for New Work

A proposed project should ordinarily satisfy most of the following:

Does it concern frontier AI evaluation, assurance, standards, or governance?
Does it strengthen evidence or institutional capacity?
Does it fill a documented knowledge or coordination gap?
Is Standards Body competent to conduct or convene the work?
Can the work remain independent?
Can the output be reviewed and revised?
Does it serve a public-interest purpose?
Can authority claims remain bounded?
Does it avoid unnecessary duplication?
Does it fit the canonical architecture?
Can conflicts and security be managed?
Is the work more than public commentary?

Projects that fail the mission test should be narrowed, referred elsewhere, or declined.

40. Source-of-Truth Hierarchy

Identity and public language should follow this hierarchy:

PROJECT_IDENTITY.md
PROJECT_MANIFESTO.md
BRAND_POSITIONING.md
FOUNDATIONS.md
The eight individual foundation papers
TERMINOLOGY.md
WEBSITE_SOURCE_OF_TRUTH.md
Individual publication files
Temporary drafts and working notes

This hierarchy applies only to identity and public positioning.

Specialized documents may govern their own domain once approved.

Examples:

GOVERNANCE_FRAMEWORK.md governs formal governance.
EVIDENCE_STANDARDS.md governs evidence requirements.
TRANSPARENCY_FRAMEWORK.md governs disclosure.
STANDARDS_DEVELOPMENT_PROCESS.md governs standards development.

Where a specialized document would materially change project identity or authority, this document must also be updated.

41. Document Architecture Position

Standards Body's canonical architecture includes four layers.

Layer 1: Philosophy

Defines:

Identity
mission
first principles
foundations
theses
positioning

Layer 2: Intellectual Framework

Defines:

Writing
editorial rules
terminology
research
evidence
taxonomy
evaluation philosophy

Layer 3: Institutional Design

Defines:

Governance
standards process
accreditation concepts
contributors
transparency
partnerships
long-term development

Layer 4: Living Knowledge Base

Maintains:

Website source
sources
people
timeline
cases
failures
signals
future work
versions

The architecture should remain coherent.

New files should have a defined role and avoid duplicating canonical sources.

42. Identity Change Control

42.1 Material Identity Changes

A material change includes:

Mission change
name change
core line change
institutional status change
formal authority
legal organization change
certification or accreditation role
regulatory role
major partnership affecting independence
significant funding model change
governance transformation

42.2 Required Change Record

A material change should document:

Proposed change
reason
evidence
institutional effect
authority effect
public interpretation
conflicts
governance approval
transition
affected documents
effective date

42.3 Public Notice

Material public identity changes should be visible in:

This document
VERSION_HISTORY.md
Website source of truth
Relevant public pages
Partnership materials

42.4 No Silent Authority Expansion

Standards Body should never expand its claimed authority through gradual language changes without formal review.

43. Initial Institutional Commitments

Standards Body commits to:

Describe its present status accurately.
Distinguish research from standards, standards from certification, and certification from regulation.
Publish versioned canonical documents.
Correct material errors.
Disclose material conflicts and funding according to approved frameworks.
Preserve reasoned dissent.
Avoid unsupported claims of consensus.
Avoid implying government or international authority.
Keep public claims within evidence.
Treat contributor credit seriously.
Support participation beyond dominant institutions.
Protect sensitive information proportionately.
Review standards and frameworks for capture and market effects.
Maintain a public record of superseded and withdrawn work.
Evaluate its own institutional development before claiming new roles.

44. Identity Success Criteria

Standards Body's identity is functioning well when:

A reasonable reader understands what the project is
A reasonable reader understands what it is not
Public materials do not imply authority the project lacks
Researchers see credible intellectual depth
Developers see a serious but independent institution
Evaluators see useful infrastructure rather than a competitor claiming premature authority
Governments can use the work without mistaking it for official policy
Open-source communities see meaningful participation pathways
Public-interest actors see accountability and inclusion
Publications remain consistent
Partners cannot use the name for unsupported endorsement
Institutional growth remains deliberate
Corrections increase rather than reduce trust
The project becomes known for evidence, precision, and institutional integrity

45. Identity Failure Conditions

The project should pause and review its identity if:

Public audiences repeatedly believe it is a regulator
Partners use its name as certification
Funding creates publication control
One company or jurisdiction dominates the agenda
Documents make inconsistent authority claims
The project begins issuing broad safety judgments
Prestige exceeds competence
Institutional expansion outruns governance
Standards are published without testing or process
Contributors cannot challenge leadership
Corrections are hidden
The project becomes primarily promotional
The project becomes primarily ideological
The public cannot identify who is accountable

46. Canonical Public Position

The following position governs all public descriptions:

Standards Body is an independent research and institutional-design project developing foundations for frontier AI evaluation, assurance, standards, and governance. Its present work consists of research, framework development, shared terminology, institutional design, and public knowledge infrastructure. Standards Body is not currently a regulator, accreditation body, certification body, or governmental authority. Its long-term purpose is to help make consequential frontier AI claims more credible, current, independently reviewable, and internationally interpretable.

47. Final Identity Statement

Standards Body exists because frontier AI requires more than technical progress and more than broad policy principles.

It requires institutions capable of producing and interpreting evidence.

Those institutions need:

Clear terms
valid methods
protected tests
independent review
competent evaluators
proportionate requirements
aligned incentives
international interoperability
transparent governance
correction

The project begins before formal authority.

It begins with foundations.

Standards Body should be judged not by the authority implied by its name, but by the quality of the intellectual and institutional infrastructure it builds.

Its identity is therefore deliberately bounded.

It is ambitious about the problem.

It is careful about its present role.

It supports beneficial progress.

It takes high-consequence uncertainty seriously.

It seeks shared understanding without forced uniformity.

It values contribution over prestige.

It treats standards as institutions that must earn trust.

The defining commitment of Standards Body is this:

Build the foundations first. Make the evidence credible. Keep authority accountable.

Revision Record

Version 1.0

Date: July 16, 2026

Change type: Complete foundational edition

Summary: Establishes the canonical identity of Standards Body, including its mission, vision, current institutional stage, authority boundaries, public-interest orientation, core principles, audiences, work areas, foundational library, relationships with governments and industry, independence, political and ideological position, voice and tone, naming, approved public descriptions, disclaimers, prohibited claims, contributor and partnership identity, funding boundaries, public trust position, website and publication requirements, institutional identity tests, public FAQ, media responses, risks, mission test, source-of-truth hierarchy, change control, initial commitments, success criteria, and final public position.

Status: Approved foundational source.