The theme of collaboration surfaced again in the session on The Use of AI for Model Risk Management. It was clear from the discussion that financial institutions are increasingly applying artificial intelligence to measure credit, anti-money-laundering (AML) and market risk but deploying AI carries new unquantifiable risks like reputation risk. The panelist each gave their perspectives on governance challenges when using AI, such as bias, transparency, interoperability and privacy. The panel agreed that fostering a collaborative environment for FIs and regulators to solve these challenges is essential. However, Manuel Morales, Chief AI Scientist at National Bank of Canada added that “academics have done a lot of research in this area” and that “academics, data scientists and other external parties should be brought into the discussion.”
There seemed to be no straightforward answer to the question posed in the title of the session: Is RegTech the Answer to Canada’s Money Laundering Woes? All panelists agreed that due to the scale of the money laundering problem in Canada, technology is needed but it is not the only answer. Organizations are often not agile enough to be proactive and to be successful, technology change needs to be part of an overall strategy and have support from senior leaders. A strong risk culture and a solid understanding of the business is important but also financial institutions must strive to ‘know their data’. The conclusion to this discussion was that a more unified approach to data, such as a movement towards data standardization, would benefit the entire ecosystem.
In the afternoon, the conversation moved to the topic of derivatives compliance and the progress that has been made in market supervision over the last few years. Julie Rochette, TMX and Lafleche Montreuil, Desjardins spoke about how FINRA has moved towards a more evidence-based approach to supervision by showing regulated entities examples of ‘this is what I see’ which leads into a two-way dialogue rather than the old approach of regulator coming in, and saying “you’re doing this wrong.” Julie Rochette said she “hoped to see a move away from a sequential rules-based approach towards the use of AI to detect the unusual and/or market manipulation or system-gaming by a market participant or trader.”
We left the real fun to the end! Sagar Aggarwal, Head of Research at Fintech Cadence, presented a ‘Data Blitz’ and took us through three trends that are making data protection challenging: 1) numerous and often overlapping verticals, 2) added complexity due to the increased use of third parties by FIs and 3) the escalating sophistication of criminals and attacks. Sagar then introduced a few case studies from their program including:
For more information on these firms, please contact firstname.lastname@example.org.
Written by: Donna Bales, Co-founder and Member of the Board, CRTA