Canadian Guardrails for Generative AI – Code of Practice

• Authority: Government of Canada
• Jurisdiction: Other countries
• Collections: Miscellaneous documents
• Most recent activity: Enacted
• Most recent activity date: 2023-08-16

Risk subdomain coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the risk subdomain are explicitly described.

• Unfair discrimination and misrepresentation
• Unequal performance across groups
• AI system security vulnerabilities
• Disinformation, surveillance, and influence at scale
• Cyberattacks, weapon development or use
• Fraud, scams, and targeted manipulation
• Overreliance and unsafe use
• Loss of human agency and autonomy
• Governance failure
• AI possessing dangerous capabilities
• Lack of capability or robustness
• Lack of transparency or interpretability

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the risk subdomain are described.

• Exposure to toxic content
• False or misleading information
• Pollution of information ecosystem
• Competitive dynamics
• AI pursuing its own goals

All other risk subdomains had no coverage in this document.

Risk Coverage Summary

The document has good coverage of approximately 12-14 subdomains, with strong focus on malicious actors (4.1, 4.2, 4.3), AI system security (2.2), discrimination and bias (1.1, 1.3), misinformation (3.1), human oversight (5.1, 5.2), governance (6.5), and AI safety failures (7.1, 7.2, 7.3, 7.4). Coverage is concentrated in security, misuse prevention, fairness, and AI safety domains.

Sector coverage

Good coverage — Comprehensive governance measures or mitigations with clear procedures relating to the sector are explicitly described.

• Information

Minimal coverage — Brief mention but no specific mitigations, recommendations or governance measures relating to the sector are described.

• Scientific Research & Development Services
• Health Care & Social Assistance

All other sectors had no coverage in this document.

Sectors Summary

This is a cross-sectoral voluntary code of practice that applies to all developers, deployers, and operators of generative AI systems in Canada, regardless of sector. The document does not limit its scope to specific industries but provides examples of misuse in healthcare and legal contexts.

Legislative Status

Soft Law

Legislative Status Summary

This is a voluntary code of practice proposed by the Government of Canada to be implemented ahead of binding regulation. The document explicitly states it will be 'implemented on a voluntary basis by Canadian firms' and uses predominantly voluntary language throughout.

Actors

Proposer

• Entity Types: AI_Governance_Actor
• Entities: Government of Canada; AI Advisory Council
• Summary: The Government of Canada is the clear proposer of this code of practice, as stated throughout the document. The AI Advisory Council is mentioned as providing expert review during the engagement process.

Target

• Entity Types: AI_Developer, AI_Deployer
• Entities: Canadian firms
• Summary: The code explicitly targets developers, deployers, and operators of generative AI systems, with specific commitments outlined for each category throughout the document. It is intended for voluntary implementation by Canadian firms.

Enforcer

• Summary: No enforcement body or enforcement mechanisms are specified in this voluntary code of practice. The document is explicitly voluntary and does not establish enforcement authority.

Monitor

• Entity Types: AI_Governance_Actor
• Entities: Government of Canada
• Summary: While no formal monitoring body is explicitly designated for this voluntary code, the Government of Canada is implicitly positioned as the monitoring entity through its engagement process and future regulatory role under AIDA.

AI Lifecycle Stages

Lifecycle stages covered by the documents, as defined in the OECD Framework for the Classification of AI Systems

• Plan and Design
• Collect and Process Data
• Build and Use Model
• Verify and Validate
• Deploy
• Operate and Monitor

AI Lifecycle Summary

The document comprehensively covers all stages of the AI lifecycle, from planning and design through deployment and ongoing monitoring. It emphasizes safety considerations throughout the system's lifecycle and includes specific requirements for data curation, model development, testing, deployment oversight, and post-deployment monitoring.

Technical Scope

Terms relating to technical scope that are mentioned or implicitly referenced in the document

• AI Models
• AI Systems
• Generative AI

Technical Scope Summary

The document explicitly focuses on generative AI systems throughout, with multiple references to their distinguishing features and capabilities. It does not explicitly mention frontier AI, general purpose AI, foundation models, or specific compute thresholds, but does reference the broad capabilities and scale of generative AI systems.