Certify Web Content Certify Web Content Certify Web Content
Bologna, Italy
(from 8 to 22)
Certify Web Content Certify Web Content Certify Web Content

The </AI> Protocol | Human-Supervised AI Framework & Standard

A new symbol is born: </AI>

It means something simple:

AI can support. Responsibility, review and publishing are always human.

The </AI> Protocol is a human-supervised AI framework and standard designed to make that responsibility explicit, structured and verifiable.

More and more content online is being published by AI without real human supervision. Posts. Comments. Articles. Messages. Automatic. Not really read, not really understood. Just published.

With the rise of autonomous agents, this dynamic will only accelerate.

We work every day with digital evidence, identity abuse, deepfakes, impersonation cases and real reputational damage. These are not abstract problems. These are real people, real consequences.

And we know what happens when responsibility becomes “no one’s responsibility”.


Access the EVIDE evidentiary registry

Explore the protocol infrastructure:

→ The </AI> Protocol
→ From decision to defensible structure: when supervision becomes evidence
→ Read the Public Technical Specification
→ Verify a CWC Code in the Public Registry
→ EVIDE – Evidentiary registry for digital content and decisions
→ CWC Registry Policy
→ Request an Official CWC Verification Code
→ AI Governance Documentation Framework
→ Implementation Guide: verifiable AI supervision
→ Oversight Bias: why human supervision can fail in AI systems
→ Decision Attestation Layer: the missing evidentiary layer in AI governance
→ AI Evidence Officer: proving human supervision in artificial intelligence systems
→ Evidentiary Layer in AI Governance
→ </AI> Protocol FAQ: questions and answers about the framework

Related reading:

→ AI Data Poisoning: The Attack No Antivirus Can Stop
→ Human in the loop: why saying there is human oversight is not enough
→ Real cases: when AI governance fails — and what should have been provable
→ AI Governance: when something has already gone wrong – forensic reconstruction and digital evidence

Work with us:

→ Legal Partners Network

Understanding the </AI> Protocol

If this is your first time here, this is a quick orientation.

The </AI> Protocol is not a single page. It is a structured system made of distinct layers.

In short:

Governance defines how decisions should happen.
The </AI> Protocol makes those decisions provable.

Why this matters for organizations

Most companies say they use human supervision in AI workflows.

But when something goes wrong, an audit, a dispute, a misleading output, a reputational incident, the real question becomes very simple:

Who was actually responsible for reviewing that output?

“Isn’t a system log enough?”
A system log is a technical self-declaration.
The </AI> Protocol is forensic-level evidence (L3), anchored to a verified identity and a public registry.
In a legal context, the difference is between something that can be questioned and something that can be verified.

Without a structured evidentiary framework, responsibility tends to remain vague. It spreads across the organization, across teams, across policies, across generic approval chains. In practice, that often means no one can clearly demonstrate who reviewed what, when, and under which conditions.

The </AI> Protocol changes that logic.

Protocol Scope

The </AI> Protocol does not define who holds authority or how decision-making processes should be structured upstream.

These aspects belong to governance models, organizational policies, and decision-making frameworks.

The protocol operates downstream.

It does not validate authority. It makes events verifiable.

Regardless of how governance is defined, the framework enables organizations to demonstrate, in a structured and defensible way, what was actually reviewed, by whom, when, and under which conditions.

It creates a model in which supervision is no longer a generic organizational statement. It becomes a documented, attributable and verifiable act tied to a specific human identity.

That matters for one reason above all:

when responsibility becomes attributable, supervision becomes real.

A named human supervisor who knows that their review is identifiable, timestamped, integrity-protected and externally verifiable will review with a different level of care than a system where “human oversight” exists only as a policy phrase.

This is the structural value of the framework:

  • it reduces distributed and ambiguous liability
  • it makes review actions attributable to a real person
  • it creates evidence that can survive audit, dispute and legal scrutiny
  • it turns human supervision from theory into accountability

In other words, the protocol does not only help organizations say that a human was involved.

It helps them prove who was responsible.

Turn AI governance into an evidentiary chain

From AI output to verifiable responsibility

AI Output
↓
Human Supervisor
↓
Identity Anchor
↓
Content Hash
↓
Qualified Timestamp
↓
Public Registry Reference
↓
Verifiable Responsibility


See how evidence is recorded in EVIDE

How the </AI> Protocol (the framework) works

The </AI> Protocol is not just a visible marker placed near content. It is a complete process designed to turn human supervision into something attributable, structured, and verifiable.

At the highest level, the framework works like this:

  • a human supervisor is identified and made accountable
  • the AI-assisted output is preserved with integrity protections
  • the supervision action is recorded as a Human Oversight Event (HOE)
  • the decision is elevated through the Decision Attestation Layer
  • the verification becomes publicly referenceable through a CWC code and registry entry

In practical terms, this means that human review is no longer just declared. It becomes documented, timestamped, attributable, and independently verifiable.

Framework flow

1. Human Supervisor
A real person takes responsibility for reviewing the output.		 2. Identity Anchoring
The supervisor is linked to a verifiable identity baseline via DAPI.		 3. Output Integrity
The reviewed content is protected through SHA-256 hashing and timestamping.
4. Human Oversight Event
The supervision action is recorded as a verifiable, timestamped event.		 5. Decision Attestation
The review becomes structured evidence, not just a system log.		 6. Public Verification
A CWC code and registry entry allow independent third-party verification.

"The

What organizations need to follow

Organizations that want to move from generic AI use to verifiable, defensible supervision should follow a clear order of implementation. The sequence is simple:

  • Step 1 — Define the Human Supervisor role: someone must be clearly responsible for reviewing AI-assisted output or decisions.
  • Step 2 — Protect the reviewed output: the exact content reviewed must be preserved with integrity protections, so it cannot be silently changed later.
  • Step 3 — Record the supervision action: every meaningful review should become a Human Oversight Event with identity, time, decision, and context.
  • Step 4 — Formalize governance rules: the organization should define how approval, escalation, rejection, and accountability work.
  • Step 5 — Activate public verification: when appropriate, the declaration should be anchored through a CWC code and public registry reference.
  • Step 6 — Be ready for audit and dispute: the whole point is not only internal governance, but the ability to demonstrate what happened if challenged later.


Access the EVIDE evidentiary registry

Without these steps, human supervision often remains a policy statement. With them, it becomes verifiable evidence.

Why we started using </AI>

That’s why we decided to start signing our content with: </AI>

It means only one thing:

  • AI can assist.
  • Responsibility remains human.
  • Every content is reviewed before publishing.

Every post we publish is opened, read, checked and consciously approved by a human. With all the limits that implies. Yes, even with mistakes: we are human!

This is not anti-AI

This is not anti-technology. We use AI as a tool. AI can accelerate, suggest, summarize. AI cannot take responsibility because it cannot understand human consequences.

What this means in practice

When you see </AI> on our content, it means:

  • a real person reviewed it
  • a human decided to publish it
  • there is accountability behind it

No autonomous publishing. No unattended output. Just human responsibility.

Examples

Bio

</AI> = Human-supervised content. No autonomous publishing.

Social explanation

</AI> = AI can assist with processes, but responsibility, review, and publication remain human.

Post signature

Certify Web Content
</AI> Human supervision guaranteed

Copy & paste snippet

If you want to adopt </AI> on your website or blog, you can copy and paste one of these snippets.

Minimal

<span title="Human-supervised content. No autonomous publishing."></AI></span>

With a short explanation

<span title="AI can support. Responsibility, review and publishing are always human."></AI></span>

Footer style

<p>
  <strong>&lt;/AI&gt;</strong>
  Human-supervised content. No autonomous publishing. Official verification code:
  <a href="https://www.certifywebcontent.com/supervised-ai/registry/" target="_blank" rel="noopener">
    CWC-XXXXXXX
  </a>
</p>

Meta Tag Example

<meta name="ai-supervision" content="AI-assisted, human-reviewed and consciously published under the </AI> Protocol." />

JSON-LD Example (Article)

{
  "@context": "https://schema.org",
  "@type": "Article",
  "creativeWorkStatus": "Human Reviewed",
  "additionalProperty": {
    "@type": "PropertyValue",
    "name": "AI Usage",
    "value": "AI-assisted, human-reviewed and consciously published under the </AI> Protocol."
  }
}

Feel free to use it. This is not a brand. It is a responsibility marker.

Why this matters

In the coming months we will see fully automated blogs, AI-driven social accounts, and endless generated discussions — without clear responsibility. This symbol is a small gesture. But this gesture speaks clearly.

For legal and compliance professionals

The </AI> marker is only the visible part of the framework. Behind it, there is a broader evidentiary architecture designed to support organizations that need to demonstrate — not just declare — that human supervision over AI systems actually occurred.

This is directly relevant for:

  • law firms assisting clients exposed to AI governance and litigation risk
  • privacy and compliance advisors building AI governance pathways
  • organizations preparing for EU AI Act obligations
  • audit, dispute, and regulatory response scenarios

The first operational entry point is the AI Oversight Risk Check: a practical evidentiary gap analysis that answers the question. If a dispute or audit were to arise today, what could your organization actually demonstrate regarding its AI oversight?

Explore the partnership framework:

→ Legal Partners Network — how law firms and compliance consultants can integrate the technical evidentiary layer into their work

For direct contact: [email protected]

FAQ

Is the </AI> Protocol anti-AI?

No. It supports AI as a tool, but requires direct human review and responsibility before publication.

How do I verify a CWC code?

The CWC code is named after our “Certify Web Content” website. Each code links to a specific URL that you can check in the public registry: certifywebcontent.com/supervised-ai/registry/

Is this a certification or a brand?

The </AI> protocol is both an open accountability marker and a public verification mechanism, not a trademark claim.

Who can adopt it?

Any author, publisher, or organization can adopt it, provided the content is genuinely reviewed and published by a human.

Is the </AI> Protocol the same as CWC?

No. </AI> is the public marker. CWC is the verification code issued by the registry. You can use the marker without a CWC code, but only registered codes carry full public verification status.

How to implement </AI> on your website

Adopting the </AI> marker can be done at different levels, depending on the degree of public verification you wish to obtain.

1. Add the visible marker

Place the </AI> marker in your footer, author bio, or near the content you publish. It signals that AI assistance may have been used, but that a human reviewed and approved the final version before publication.

2. Use the marker responsibly

The </AI> symbol is a declaration of accountability. It must reflect genuine human supervision. It is not a decorative or marketing element.

3. Request an official CWC verification code

If you want your declaration to be publicly verifiable, request an official CWC verification code via the official request page: Request a CWC verification code.

Each issued code is:

  • Issued individually
  • Linked to a specific domain or content
  • Published in the official registry certifywebcontent.com
  • Subject to minimum validation review

Only codes listed in the public registry are considered valid.

4. Include the official verification sentence

Once a CWC code has been issued and published in the registry, include the following sentence in your content:

Official verification code: CWC-YYYYMMDD-XXXX.
Codes are listed and verifiable at: https://www.certifywebcontent.com/supervised-ai/registry/

5. Add structured metadata (Optional)

You may include structured metadata (Schema.org) referencing your issued CWC code to allow automated systems to detect and verify the declaration. See the full snippet in the Public Specification.

Adopt it if you want

This is not marketing. It is a position. If you want to adopt it, do it freely. Technology is powerful.

But responsibility remains human.

Kind regards,


Explore the EVIDE evidentiary registry

</AI> Emanuel Celano
Founder, Certify Web Content

The framework is publicly defined as “The </AI> Protocol” and is forensically certified through CertifyWebContent.
This documentation constitutes a verifiable, timestamped record of its structure, concepts, and implementation.
Share
Copy Link
×