Quebec Financial Markets Regulator Releases Draft AI Guidelines For Comment

On July 3, 2025, Quebec’s Autorité des marchés financiers (AMF) released new draft guidelines (the Guidelines) on the use of artificial intelligence (AI) in the financial sector (available in French and English). The Guidelines set out expectations for the risk-based classification of artificial intelligence systems (AIS), the lifecycle of AISs, organizational governance, risk management, and the fair treatment of clients.

The Guidelines define an AIS as a “machine-based system that, for explicit or implicit objectives, infers, from the input it receives, how to generate outputs such as predictions, content, recommendations, or decisions that can influence physical or virtual environments.”

All financial institutions should assign a risk rating to all AISs which will determine the necessary scope of all related policies, processes, and procedures. AISs with higher risk ratings require more frequent monitoring and updating. This risk rating should be reviewed annually to ensure each AIS is still properly assessed.

Institutions are expected to develop, document, approve, and implement processes and controls for each stage of an AIS’s lifecycle proportional to its risk rating. This obligation begins when an AIS is first developed or acquired and continues throughout the training, approval, use, and monitoring for the remaining lifespan of the AIS. Institutions must also have defined roles and responsibilities for their board of directors, senior management, and other stakeholders involved in the development and implementation of AISs.

Financial institutions should base AIS policies, processes, and procedures on the nature, size, and complexity of the institution’s activities, risk profile, and risk ratings assigned to each AIS. Institutions should have an inventory of all AISs and implement an internal reporting system to ensure key stakeholders regularly receive an assessment of the risks associated with AIS use across the institution.

The AMF also expects institutions to maintain high standards of ethics and integrity in the context of its use of AISs. Factors such as discrimination, bias, and low-quality data must be corrected when identified and clients should be informed when they are engaging with an AIS or receive a decision made by an AIS.

The Guidelines are currently open for consultation. Stakeholders may submit comments until November 7, 2025.

Summary By: Amy Ariganello