WORKING DRAFT · June 2026 W3C AI Knowledge Representation Community Group (AI KR CG) · structured by the Knowledge-Domains taxonomy
Across the eight knowledge domains, the majority remain a gap or only partially served; standards activity is concentrated in communication/identity, commerce, security, and regulation.
| Knowledge domain | Coverage | Summary |
|---|---|---|
| 1. Agent types & autonomy | GAP | No ratified standard describes agent types or autonomy levels; only governance frameworks reference risk tiers. |
| 2. Cognition & capability | GAP | Planning, goal specification and perception have no standards; agent memory now has an emerging W3C community group. |
| 3. Interaction, communication & identity | PARTIAL | A vibrant de-facto protocol layer and maturing identity primitives, but no single ratified web standard yet. |
| 4. Commerce & settlement | PARTIAL | Several competing de-facto payment protocols; no ratified settlement standard. |
| 5. Oversight, safety & interpretability | PARTIAL | Strong de-facto security frameworks and regulatory anchors exist; no ratified oversight vocabulary, and interpretability remains research. |
| 6. Regulation & governance | PARTIAL | Binding regulation exists (EU AI Act, DORA, MiCA); agentic-specific governance is still forming. |
| 7. Agentic standards & standards bodies | ACTIVE | Many standards bodies are now actively engaged — the landscape is filling, mostly with community groups and de-facto protocols. |
| 8. Applied domains | GAP | No agentic-AI-specific standards; applied work relies on each field's existing domain standards. |
No ratified standard describes agent types or autonomy levels; only governance frameworks reference risk tiers.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| (none dedicated) | No standard for agent type / autonomy-level description | GAP | — | — |
| EU AI Act risk tiers | Risk classification referenced for autonomy | Regulation | open | Reg (EU) 2024/1689 |
| NIST CAISI AI Agent Standards Initiative | US program; may produce autonomy/agent descriptors | Initiative | open | NIST |
Planning, goal specification and perception have no standards; agent memory now has an emerging W3C community group.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| W3C AI Agent Memory Interoperability CG | Portable agent memory specification | Proposed (18 May 2026) | open | W3C |
| Planning / goal specification | No dedicated standards venue | GAP | — | — |
| Perception | No dedicated standards venue | GAP | — | — |
A vibrant de-facto protocol layer and maturing identity primitives, but no single ratified web standard yet.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| A2A (Agent2Agent) | Inter-agent communication / coordination | De-facto | open | Linux Foundation |
| MCP (Model Context Protocol) | Agent-to-tool connectivity | De-facto | open | Anthropic / LF |
| ACP (Agent Communication Protocol) | REST-native inter-agent messaging | De-facto | open | AGNTCY / LF |
| ANP (Agent Network Protocol) | DID-based decentralised discovery | White Paper | open | W3C |
| OASF (Open Agent Schema Framework) | Agent capability metadata schema | Draft | open | AGNTCY |
| NANDA | Global agent discovery / registry | Research | open | MIT Media Lab (Project NANDA) |
| DID (Decentralized Identifiers) | Agent identity | Candidate Rec (Mar 2026) | open | W3C |
| Verifiable Credentials 2.0 | Agent credentials | Recommendation | open | W3C |
| W3C AI Agent Protocol CG | Inter-agent comms / identity / metadata | Pre-spec | open | W3C |
Several competing de-facto payment protocols; no ratified settlement standard.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| x402 | HTTP-402 agent/stablecoin payments | De-facto | open | x402 Foundation / Coinbase |
| AP2 (Agent Payments Protocol) | Mandate-based agent payment authorisation | De-facto | open | Google / Coinbase |
| ACP (Agentic Commerce Protocol) | Agent-initiated checkout | De-facto | partly gated | Stripe / OpenAI / Meta |
| ERC-8004 (Trustless Agents) | On-chain agent identity / reputation | Draft EIP | open | Ethereum |
Strong de-facto security frameworks and regulatory anchors exist; no ratified oversight vocabulary, and interpretability remains research.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| OWASP Agentic Security Initiative + Top 10 | Agentic threat taxonomy and controls | De-facto | open | OWASP |
| CSA MAESTRO | Seven-layer agentic threat model | De-facto | open | Cloud Security Alliance |
| Five Eyes joint guidance | Secure agentic-AI deployment | Guidance | open | CISA et al. |
| EU AI Act Art. 14 / prEN 18283 | Human oversight | Regulation / draft std | open | EU / CEN-CENELEC |
| Interpretability | No standard; active research | GAP | — | — |
Binding regulation exists (EU AI Act, DORA, MiCA); agentic-specific governance is still forming.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| EU AI Act (incl. GPAI) | Horizontal AI regulation | In force | open | Reg (EU) 2024/1689 |
| DORA | Financial-sector operational resilience | In force | open | Reg (EU) 2022/2554 |
| MiCA | Crypto-asset markets | In force | open | Reg (EU) 2023/1114 |
| GENIUS Act | US stablecoin statute | Statute | open | Public Law 119-27 |
| NIST CAISI AI Agent Standards Initiative | US agent standards program | Initiative (Feb 2026) | open | NIST |
| IMDA Model AI Governance Framework for Agentic AI | Singapore agentic governance | Framework | open | Singapore IMDA |
Many standards bodies are now actively engaged — the landscape is filling, mostly with community groups and de-facto protocols.
| Standard / effort | What it covers | Coverage | Openness | Source |
|---|---|---|---|---|
| W3C AI-KR CG | Knowledge representation for AI (this group) | CG | open | W3C |
| W3C AI Agent Protocol CG | Agent communication / identity | CG (pre-spec) | open | W3C |
| W3C AI Agent Memory Interoperability CG | Agent memory | CG (proposed) | open | W3C |
| Linux Foundation Agentic AI Foundation | Stewards MCP and A2A | Foundation | open | Linux Foundation |
| AGNTCY collective | Internet of Agents (ACP, OASF) | Collective | open | AGNTCY |
| C2PA | Content provenance (agent-generated content) | Standard (v2.3) | open | C2PA |
| CEN-CENELEC JTC 21 | EU harmonised AI standards (prEN 182xx) | Draft standards | open | CEN-CENELEC |
No agentic-AI-specific standards exist for this domain; agents operating in a given field reuse that field's own domain standards. There is nothing agent-specific to inventory here — this category is retained only so the eight-domain spine stays complete.
Where the inventory shows a gap, these are the standards worth developing — carried over from the June 2026 gap analysis and tagged by knowledge domain. The status column notes the state of any effort as of June 2026.
| Work item | Domain | Addresses | Status of effort (Jun 2026) |
|---|---|---|---|
| Agent Autonomy Level Vocabulary | 1 · Agent types | Reactive / deliberative / learning / goal-seeking agents | Open; governance anchors only (EU AI Act, NIST CAISI) |
| Goal Specification Schema | 2 · Cognition | Machine-readable goals & constraints | Open; no venue |
| Planning Semantics Vocabulary | 2 · Cognition | Plans, strategies, constraints | Open; no venue |
| Agent Memory Model | 2 · Cognition | Portable short / long-term / episodic memory | Active: W3C Agent Memory Interop CG |
| Agent Communication Protocol | 3 · Interaction | Web-native agent-to-agent messaging | Active (de-facto): A2A, MCP, ACP; W3C Agent Protocol CG |
| Agent Identity Specification | 3 · Interaction | Identity / credentials for non-human actors | Maturing: DID v1.1 CR, VC 2.0 Recs |
| Delegation & Authority Model | 3 · Interaction | Scoped, authorised delegation | Research: authenticated delegation |
| Human Oversight Vocabulary | 5 · Oversight | Human-in-the-loop requirements | Open; EU AI Act Art. 14 anchor |
| Creative-Agent Provenance | 5 · Oversight | Attribution of agent-generated content | Largely met: C2PA |
| Policy & Governance Annotations | 6 · Regulation | Operational / legal constraints on agents | Frameworks proliferating; ODRL alignment |
| Perception Interface Descriptors | 2 · Cognition | Sensor / input-modality description | Open; no venue |
Companion documents: the taxonomy (category spine) and the Vocabulary. Status: DRAFT — provisional, for community evaluation and maintenance. License: CC BY 4.0.