AI Summary: Australia Mandatory AI Guardrails is an proposed AI regulation
in AU
focused on high_risk_guardrails.
Maximum penalty: TBD (legislation pending).
Nerq has assessed 41,686 agents against applicable risk classifications.
AI Agent Compliance: Australia Mandatory AI Guardrails
Australia Mandatory AI Guardrails is a currently proposed regulation focused on high_risk_guardrails in AU. It establishes a risk-based framework for AI systems with penalties up to TBD (legislation pending) for non-compliance.
Overview
| Status | proposed |
| Effective Date | 2026-TBD |
| Region | AU |
| Country | AU |
| Focus Area | high_risk_guardrails |
| Max Penalty | TBD (legislation pending) |
| Per Violation | TBD |
| Source | Official text |
Risk Model
risk_based
Risk Classes
- high_risk
- general_purpose
- low_risk
High-Risk Criteria
- {'consequential_decisions': ['employment', 'financial', 'healthcare', 'government', 'legal'], 'triggers': ['significant_harm_potential', 'vulnerable_populations']}
Requirements
- Mandatory guardrails for high-risk AI
- Transparency and explainability
- Testing and monitoring requirements
- Human oversight mechanisms
Agent Risk Distribution
41,079
minimal
400
high
207
limited
Top Agents by Compliance Score
| Agent | Compliance | Risk Class | Trust Score |
|---|---|---|---|
| livekit/agents-js | 100.0 | minimal | 82.8 |
| dyoshikawa/rulesync | 100.0 | minimal | 81.3 |
| williamzujkowski/strudel-mcp-server | 100.0 | minimal | 80.3 |
| QwenLM/qwen-code | 100.0 | minimal | 80.1 |
| microsoft/azure-devops-mcp | 100.0 | minimal | 79.7 |
| opal | 100.0 | minimal | 79.6 |
| GoogleCloudPlatform/agent-starter-pack | 100.0 | minimal | 79.5 |
| laravel/mcp | 100.0 | minimal | 79.4 |
| agentgateway/agentgateway | 100.0 | minimal | 79.2 |
| RooCodeInc/Roo-Code | 100.0 | minimal | 78.9 |
Frequently Asked Questions
What are the compliance requirements under Australia Mandatory AI Guardrails?
Australia Mandatory AI Guardrails requires AI systems to meet specific regulatory standards focused on high_risk_guardrails. Nerq automatically checks AI agents against these requirements.
How does Australia Mandatory AI Guardrails classify AI risk?
Australia Mandatory AI Guardrails uses a risk-based classification: risk_based. Nerq maps each agent to the applicable risk class.
What are the penalties under Australia Mandatory AI Guardrails?
Non-compliance with Australia Mandatory AI Guardrails can result in penalties up to TBD (legislation pending). Use Nerq to identify compliance gaps before enforcement.