Meta-Layer Initiative

Comoros karimattoumanimohamed@gmail.com

Meta-Layer Initiative

USA daveed@bridgit.io

Applications and Real-Time GEN-Dispatch Meta-layer annotation provenance governance overlay presence This document introduces the concept of a Meta-layer: a programmable coordination substrate that operates above content layers on the Internet. The Meta-layer enables communities, individuals, and agents to appear, annotate, and govern together in shared digital space, independent of underlying platforms. It is not a replacement for existing web or transport protocols, but a complementary infrastructure that integrates with them. The draft outlines the motivation, terminology, use cases, implementation model, risks, security considerations, and potential IANA registries for future work. 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." Copyright (c) 2025 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.

The Internet has evolved from a document-sharing network into a global application substrate. However, it has never included a shared layer for presence, annotation, provenance, and contextual governance across domains. These functions remain fragmented, implemented in proprietary platforms or plugins, without interoperability or transparency. The idea of a higher-level coordination or annotation layer above content is not new:

Vannevar Bush (1945)

“As We May Think,” The Atlantic Monthly, introduced associative trails—linked paths of thought that presaged hypertext and the idea of connecting knowledge above documents.

Ted Nelson (1965)

“Complex information processing: a file structure for the complex, the changing and the indeterminate” introduced hypertext as a precursor to a cross-page meta-layer.

Douglas Engelbart (1968)

“The Mother of All Demos” (FJCC, San Francisco) publicly demonstrated NLS with hypertext, on-screen overlays, and view controls—effectively a layer above documents.

Tim Berners-Lee (2001)

“The Semantic Web,” Scientific American, framed a data/meaning layer on top of the Web, i.e., machine-understandable metadata layered over pages.

Marc Andreessen (2012)

“Why Andreessen Horowitz Is Investing in Rap Genius,” describing widespread web annotation as the “missing layer of the Internet.”

The Meta-layer Initiative seeks to translate these longstanding conceptual foundations into open, interoperable infrastructure under IETF stewardship—turning decades of vision into a standard that integrates presence, annotation, provenance, and governance as native Internet functions.

Current IETF protocols provide robust foundations for transport (TCP, QUIC), security (TLS), and identity (OAuth, OIDC, SCIM). However, the Internet still lacks standardized primitives for:

• Presence: expressing who is here, under what rules, and with what visibility.
• Annotation: attaching structured meaning (claims, challenges, polls, bridges) to content across domains.
• Provenance: cryptographically linking contributions to identity and context.
• Governance: enabling communities to compose and enforce rules transparently.
• Agent Containment: running AI and automated processes inside bounded, verifiable execution environments.

Today, these behaviors exist only as fragmented features inside proprietary platforms. This results in interoperability gaps, inconsistent privacy guarantees, lack of portability, and absence of shared governance mechanisms. While the W3C Web Annotation Data Model (2017) has defined a standard format for content-level annotations, it does not address cross-domain interoperability, provenance, or rule-based governance. The Meta-layer complements W3C’s work by proposing a protocol-level substrate—capable of operating across applications and domains—where annotations, presence, and governance can interoperate securely and transparently. The absence of such a substrate has long been recognized: the ability to annotate and govern content was described as a “missing feature” of the web browser, and calls to explore a “meta-environment above the page” have been made by early Internet pioneers. As Marc Andreessen noted in “Why Andreessen Horowitz Is Investing in Rap Genius” (2012), this “missing layer” reflects a longstanding need for interoperable annotation infrastructure. These concepts build on the architectural vision outlined in “The Metaweb: The Next Level of the Internet” (Bridgit DAO, CRC Press/Taylor & Francis, 2023), which introduced the concept of a “meta-layer above the webpage” as a civic and computational trust substrate. This draft operationalizes that vision for standardization within the IETF context.

This work aligns with several ongoing activities across IETF Areas and external web-standard bodies.

Defines application-layer primitives for presence, annotation, and overlays, complementing ongoing work such as MIMI (Messaging Interoperability) and HTTP APIs. The Meta-layer’s semantic and contextual overlay model complements W3C’s Web Annotation work by introducing interoperable signaling, provenance, and governance primitives at the Internet protocol layer.

The Meta-layer depends on secure identity, accountability, cryptographic provenance, and trusted execution environments (TEEs). It builds upon and extends existing work in OAuth, OIDC, Privacy Pass, and SCIM for federated identity and access control; RATS (Remote ATtestation ProcedureS) and EAT (Entity Attestation Token) for verifying trustworthiness of execution environments; COSE (CBOR Object Signing and Encryption) and CFRG for cryptographic signing and post-quantum resilience; and SUIT (Software Updates for IoT) for maintaining verified code integrity within TEEs. TEEs are thus positioned as security primitives within the IETF SEC Area, ensuring that agents in the Meta-layer execute in verifiable, policy-constrained, and auditable contexts.

The governance and AI-containment aspects of the Meta-layer overlap with ongoing research in PEARG (Privacy Enhancements and Assessments RG) and RASPRG (Research and Analysis of Standard-Setting Processes RG). The initiative can also contribute to IRTF and IAB workshops on AI accountability, provenance, and sustainable governance models.

Since the Meta-layer crosses multiple areas (ART, SEC, OPS, IRTF), GEN-Dispatch is an appropriate venue to discuss scope and determine whether a dedicated Working Group (WG) or Research Group (RG) is warranted.

The Meta-layer aims to be complementary to ongoing efforts in W3C (e.g., Web Annotation, ActivityPub, and provenance standards) and ISO/IEC JTC1 AI frameworks, by providing a network-layer and governance substrate that ensures interoperability, accountability, and trust across ecosystems.

(Working definitions)

Overlay

Semantic/visual layer rendered above digital content, carrying presence indicators, tags, and interactions; governed by community rule modules.

Smart Tag

Typed, structured annotation (e.g., note, claim, challenge, poll, bridge); signed, timestamped, interactive, filterable.

Bridge

Semantic link connecting two pieces of content (support, challenge, context) across domains.

Presence

Identity expression in digital space, scoped by context and rules (visible, pseudonymous, invisible).

Governance Module

Composable logic defining rules for interaction, moderation, participation, and policy enforcement within an overlay.

Agent

Automated or semi-autonomous process (AI, bot, scripted service) operating within the Meta-layer under policy constraints.

TEE (Trusted Execution Environment)

Secure, attestable runtime container for agent execution, supporting constraints, logging, attestation.

Provenance

Verifiable origin, context, and authorship of tags/actions/agent behaviors via signed metadata and timestamps.

The Meta-layer operates above existing content without requiring fundamental Web changes. Functions are delivered via extensions, SDKs, and open APIs.

Lightweight extension or embeddable SDK renders overlays on existing sites. Overlays carry smart tags, presence, governance. Interoperable and governed by open registries (unlike closed annotation tools).

Sites integrate Meta-layer widgets or frames (e.g., a semantic sidebar in e-learning portals) via web-embed SDK. No browser installation required for end users in these contexts.

Identity, tags, and governance rules are portable. Provenance (signatures, timestamps) ensures authenticity across domains.

Agents operate in bounded execution environments (e.g., TEEs) with policy-defined capabilities, rates, and auditable logs.

APIs expose registries for tag types, badge schemas, governance modules. Third parties define new tag types, build overlays, or fork rule modules. Interop via stable identifiers (IANA-registered if standardized).

Early opt-in communities (e.g., research/fact-checking overlays). Later: native integrations once interop/security are proven. No "flag day"—coexists and incrementally extends today’s Internet.

Federated identity, proof-of-humanity, and contextual filters enable communities to restrict participation (e.g., verified humans; scoped agent permissions) and create bot-resistant, trustable interaction zones.

Smart tags and bridges make annotations portable and filterable across sites, building shared knowledge graphs with provenance.

Agents run in attested TEEs with logged behaviors and community-defined permissions—preventing unbounded automation and interaction while enabling useful collaboration.

• Identity fraud and bots: federated identity, contextual privileges, proof-of-humanity (when needed).
• Governance capture: modular rule modules, open registries, and forkable governance.
• Annotation spam or overload: attention-based rendering, overlay moderation, reputation weighting, rate limits.
• Privacy loss: scoped presence, pseudonymity, user-controlled visibility, data-minimizing defaults.
• Fragmentation: shared registries for tag types, governance modules, and semantic formats, with room for extensions or forks.

• Identity and Authentication: leverage OAuth, OIDC, SCIM, Privacy Pass.
• Cryptographic Provenance: sign and timestamp tags and bridges; consider hybrid/post-quantum algorithms over time.
• Agent Containment: bounded TEEs or verifiable sandboxes with attestation and enforceable policy.
• Privacy: opt-in presence and annotation; scoped visibility; minimize metadata exposure.
• Registries and Extensions: standardized, auditable registries (designated expert review) to prevent identifier squatting or abuse.

No immediate IANA actions requested. If standardized, potential new registries include:

• Meta-layer Smart Tag Types Registry (e.g., note, claim, bridge, poll, challenge).
• Governance Module Registry (reusable rule modules for overlays/participation/moderation).
• Badge and Role Types Registry (moderator, validator, scribe, etc.).

Registries should balance extensibility with security and interoperability, using clear specification references and designated-expert review.

Normative References Informative References The Atlantic Monthly Scientific American

Karim ATTOUMANI MOHAMED Meta-Layer Initiative Email: karimattoumanimohamed@gmail.com Daveed Benjamin Meta-Layer Initiative Email: daveed@bridgit.io