In November 2021, the Autorité des marchés financiers (AMF) of Québec published a report (only in French) on artificial intelligence (AI) in finance, recognizing its principal functions in valuation, incentives, optimisation, advice and information.
Given AI’s growing role, the report highlights critical considerations not only on the responsible use of personal data, protection of privacy and the principle of equity, but also on the ethics of using AI in the finance sector and the challenges of governance.
From a consumer viewpoint, the main issue on the use of AI in finance concerns trust. Some of the major consumer questions include, will the AI respect individual autonomy, is the AI system intruding in and monitoring private lives, what is the risk of bias and algorithmic discrimination and finally, are the decisions ultimately made by humans?
Based on an understanding of different perspectives in the industry and from a consumer point of view, the report makes ten recommendations to encourage the responsible use of AI systems and protect consumers. The first three recommendations focus on what the regulator can do. The remaining seven focus on key actions for financial institutions (FI) , including:
Assess how AI would be relevant to the FI’s mission and client expectations, before developing and using AI
Adopt a governance framework specific to the use of AI, attributing human responsibility to decisions made by the AI system
Adapt the complaint procedures, including dispute resolution and mediation, particularly as AI decision making grows while decreasing consumer control
Facilitate the adoption of certification procedures and audit AI systems
Ensure AI systems are resilient, efficient, robust and safe, in order to financial market stability
Ensure AI systems do not undermine equity and do not reinforce discrimination
Ensure AI systems respect consumer autonomy and communicate information required for consumers to give free and informed consent