Who this is for: AI platform teams deploying agents in or serving Chinese markets, compliance officers at multinational AI companies, legal counsel advising on cross-border AI agent deployment, and GRC teams managing multi-jurisdictional compliance.

Status: In force since July 15, 2026. "Implementation Opinions on AI Agents" jointly issued by CAC, NDRC, and MIIT (May 8, 2026). World's first dedicated regulatory category for AI agents. Mandatory filing required for agents in sensitive sectors (healthcare, transportation, media, public safety). Treats agentic AI as "future digital infrastructure" -- a governance object, not just an application layer. ByteDance (Doubao) and Alibaba (Qwen) shut down non-compliant features overnight.

Contents

1. What China Requires 2. Obligation-to-Procedure Mapping 3. Detailed Procedure Cards 4. Cross-Border Considerations 5. Quick Reference 6. Quick Start 7. References

1. What China Requires

China's "Implementation Opinions on AI Agents" establishes the world's first dedicated regulatory category for autonomous AI agents. Unlike the EU AI Act (which regulates AI systems broadly) or the US approach (guidance-based), China treats AI agents as a distinct governance object requiring purpose-built controls.

Three-Tier Decision Authorization

TierScopeAuthorizationExample
Tier 1 -- RoutineLow-impact, reversible actionsAgent-autonomousInformation retrieval, formatting, scheduling
Tier 2 -- SignificantConsequential but manageable actionsHuman-confirmed before executionFinancial transactions, content publication, data modification
Tier 3 -- CriticalHigh-impact, irreversible, or safety-relevantHuman-approved with full audit trailMedical recommendations, infrastructure changes, legal filings

Additional Requirements

2. Obligation-to-Procedure Mapping

China RequirementSWT3 ProcedureWhat It WitnessesEvidence Produced
Three-tier authorizationAI-ACC.1Access scope and permission level per actionAnchor with authorization tier, scope, decision outcome
Pre-inference gateauthorization_idGate decision before executionAnchor with gate result, authorization source, tier classification
Mandatory CAC filingAI-AUDIT.1 + AI-TRANS.1Audit log integrity + transparency disclosureAnchors proving filing documentation and disclosure events
Compliance testingAI-PERF.1 + AI-ROBUST.1Performance validation + adversarial robustnessAnchors with metrics, benchmarks, perturbation survival rates
Human overrideAI-HITL.1Human intervention decision and rationaleAnchor with decision type, rationale, authority level
Emergency haltAI-EMRG.1Emergency lifecycle (halt, contain, recover)Anchor with emergency type, actions halted, recovery plan
Agent identity registrationAI-ID.1Agent identity bound to every inferenceAnchor with agent_id, identity persistence verification
Tool access controlAI-TOOL.1Tool invocation with authorization contextAnchor with tool name, invocation context, authorization status
Content governanceAI-GRD.1Guardrail enforcement stateAnchor with guardrail type, trigger status, bypass attempts
Delegation and chainingAI-DEL.1Agent-to-agent delegation with constraintsAnchor with delegator, delegatee, scope, constraints

3. Detailed Procedure Cards

AI-ACC.1 + authorization_id

Three-Tier Decision Authorization

China requires: A three-tier framework classifying every agent action by impact level, with escalating human involvement for higher tiers.

How SWT3 addresses it: The authorization_id pre-inference gate classifies each action before execution. wrapAccess() mints an AI-ACC.1 anchor recording the authorization tier, scope, and decision outcome. Tier 2 and Tier 3 actions include confirmation or approval metadata in the anchor.

What to show the examiner

Filter the ledger by AI-ACC.1. Verify tier distribution matches the agent's operational profile. Tier 3 actions without human approval anchors (AI-HITL.1) are compliance violations. Cross-reference authorization_id with the pre-inference gate log.

AI-HITL.1 + AI-EMRG.1

Human Override and Emergency Halt

China requires: Operators must be able to halt agent actions at any time. Emergency containment must be immediate.

How SWT3 addresses it: witnessHumanOverride() records every human intervention with decision type (approve, deny, escalate), rationale, and authority level. witnessEmergency() handles the emergency lifecycle -- halt, contain, recover -- with timestamps at each stage.

What to show the examiner

AI-EMRG.1 anchors prove all active agents were halted during emergency events. Response time from trigger to halt should be documented. AI-HITL.1 anchors should exist for all Tier 2 and Tier 3 decisions.

AI-ID.1 + AI-TOOL.1

Agent Registration and Tool Control

China requires: Mandatory filing requires each agent to have a registered identity. Tool access must be controlled and auditable.

How SWT3 addresses it: agent_id binds a persistent identity to every witness anchor minted by the agent. wrapTool() mints an AI-TOOL.1 anchor for every tool invocation, recording the tool name, invocation context, and authorization status.

What to show the examiner

Verify agent_id is consistent across all anchors for a given agent. Cross-reference tool invocations with the agent's authorized tool manifest. Unauthorized tool access attempts should be flagged by AI-GRD.1 anchors.

AI-PERF.1 + AI-ROBUST.1

Pre-Deployment Compliance Testing

China requires: Agents must pass compliance testing before deployment. Product recall provisions apply to non-compliant agents.

How SWT3 addresses it: witnessPerformance() validates the agent against declared benchmarks. witnessRobustness() tests adversarial resilience. Both produce anchors with timestamped results that prove testing occurred before the first production inference.

What to show the examiner

Verify testing anchors predate the first production inference anchor (AI-INF.1). The gap between last test and first deployment should be documented. Ongoing testing cadence proves continuous compliance.

4. Cross-Border Considerations

Organizations deploying AI agents in both China and the EU face overlapping requirements. China's three-tier authorization aligns conceptually with the EU AI Act's risk classification, though the specific obligations differ.

5. Quick Reference

Examiner QuestionWhere to Look
How do you classify agent actions by tier?AI-ACC.1 anchors with authorization tier metadata. Filter by tier to verify distribution.
Can you halt an agent immediately?AI-EMRG.1 anchors showing emergency halt events with response time from trigger to full stop.
Is the agent registered with CAC?AI-TRANS.1 anchors documenting filing disclosure events. AI-AUDIT.1 for filing documentation integrity.
Was the agent tested before deployment?AI-PERF.1 + AI-ROBUST.1 anchors with timestamps preceding the first AI-INF.1 production anchor.
How do you control which tools the agent can use?AI-TOOL.1 anchors per tool invocation. Cross-reference with AI-GRD.1 for unauthorized access attempts.
How do you handle agent-to-agent delegation?AI-DEL.1 anchors with delegator, delegatee, scope, and constraint metadata. Verify scope does not exceed the delegator's tier.

6. Quick Start

# Install the SDK
pip install swt3-ai

# Initialize with agentic profile
swt3 init --profile autonomous-systems --tenant YOUR_TENANT

# Run the demo
python -m swt3_ai.demo

# TypeScript
npm install @tenova/swt3-ai
npx swt3-init --profile autonomous-systems

Full SDK documentation: sovereign.tenova.io/docs

Create a free account: sovereign.tenova.io/signup

7. References