On December 5, 2024, the Canadian Securities Administrators (CSA) issued CSA Staff Notice and Consultation 11-348, a comprehensive guide addressing the utilisation of Artificial Intelligence (AI) systems within capital markets. This initiative comes as a response to the increasing integration of AI technologies in financial services, which the CSA acknowledges could enhance market efficiency but also presents unique challenges and risks for investors.

The Notice identifies key themes governing the responsible deployment of AI across capital markets, articulating considerations market participants should contemplate as they adopt AI solutions. In conjunction with the Notice, the CSA has launched a public consultation period, allowing stakeholders to submit their comments and perspectives on the potential regulation of AI in capital markets. This consultation period is open until March 31, 2025.

Outlined within the Notice are five central themes relevant to the use of AI in financial sectors:

  1. Technology and Securities Regulation: The CSA clarifies that existing securities laws remain technology-neutral, focusing on the activities of market participants rather than the specific technologies employed. Thus, the treatment of AI systems will depend on their functional applications.

  2. Governance and Oversight: Market participants are urged to implement governance practices tailored to AI that adequately address the distinct risks and challenges this technology may pose. A recommendation includes incorporating a “human-in-the-loop” model to monitor AI systems, ensuring they operate within intended parameters.

  3. Explainability: A call for achieving high levels of explainability in AI systems was highlighted, advocating for clarity in the system’s reasoning processes. The CSA advised against "black box" AI systems that lack transparency, stressing that capital markets thrive on disclosure and transparency.

  4. Disclosure: The CSA promotes transparency in the application of AI, advising that participants should avoid generic or deceptive disclosures. They are expected to provide substantial information on how AI is integrated, enabling investors to make well-informed decisions.

  5. Conflicts of Interest: Recognising that AI systems can introduce new conflict scenarios, the CSA emphasizes the necessity for market participants to ensure that AI outputs do not lead to decisions that favour their interests over those of their clients or investors.

The Notice further extends specific guidance tailored to market participants regarding AI integration in alignment with existing regulatory obligations:

  • Registrants: Investment firms are advised to evaluate how AI influences their registration requirements. Registered advisers and dealers should verify that AI usage does not compromise the quality of advice for their clients.

  • Non-Investment Fund Reporting Issuers: Comprehensive guidance is provided on how continuous disclosure obligations relate to AI, cautioning against misleading promotional statements about AI capabilities.

  • Marketplaces: The CSA encourages the implementation of robust controls that comply with necessary regulatory standards. These include regular AI system testing and validation procedures.

  • Clearing Agencies: Subject to rigorous operational and risk management requirements, clearing agencies are advised to conduct annual reviews and vulnerability assessments of their AI systems.

  • Trade Repositories: The CSA insists that adequate controls and robust security measures are crucial for protecting sensitive data within AI systems.

  • Designated Rating Organisations and Benchmark Administrators: These entities are encouraged to maintain a transparent approach to AI utilisation, including public disclosures concerning methodologies and operational assumptions.

While the Notice does not amend current legal obligations for market participants, the CSA is keen on gathering feedback regarding potential regulatory frameworks for AI in capital markets in the upcoming months. This initiative reflects a proactive approach to managing the evolving landscape of AI technology within investment practices, bridging innovation with regulatory diligence.

The authors express gratitude to Chloe Loblaw and Patrick Lajoie for their valuable contributions to this analysis.

Source: Noah Wire Services