The Rise of AI in Canadian Family Offices: A Closer Look

Exploring the Impact of AI on Family Offices

Peter Kenter • Canadian Family Offices

Published Jan 27, 2025

A friend once suggested an intriguing scenario: a time traveler takes a computer back to the 1940s. Imagine Sir Winston Churchill equipped with artificial intelligence to help the Allies win World War II years earlier. Recognizing AI's potential, Churchill would set the traveler to compiling his memoirs.

Today, the real-world application of artificial intelligence in investing is somewhat comparable. While many focus on AI’s potential to provide foolproof investment advice, family offices often leverage its strengths in organizing records and generating reports. They also identify investment opportunities based on specific criteria.

Family offices are interested in investing in companies that utilize AI technologies. However, they remain cautious about incorporating AI into their own operations.

Current AI Adoption Rates

A recent survey from Citi Private Bank reveals a fascinating landscape. More than half (53 percent) of global family offices seek portfolio exposure to AI technologies, while an additional 26 percent show interest. Yet, fewer than 15 percent are using AI to automate tasks, develop presentations, or make forecasts.

The optimism surrounding AI often stems from media hype and misunderstandings. Many people are not clear on the different types of artificial intelligence available.

Understanding Machine Learning and Generative AI

Machine learning, a significant branch of AI, analyzes large datasets, identifies patterns, and automates intricate tasks. It helps family offices organize records, generate portfolio reports, develop presentations, and maintain compliance.

On the other hand, generative AI powers products like ChatGPT, Copilot, and DALL-E. This technology builds upon machine learning to create text, images, and other forms of content.

Despite low overall adoption rates, certain family offices and advisors that embrace AI technology express enthusiasm about its impact.

Case Study: Bell Kearns & Associates Ltd.

Bell Kearns & Associates Ltd., based in Toronto, is a multi-family office that recently collaborated with the Vector Institute. This non-profit organization aims to accelerate AI's use in businesses. In their partnership, Bell Kearns worked with master's students in computer engineering focusing on generative AI.

Helen Kearns, CEO of Bell Kearns, explains, “Though we only have 45 clients, we have over 200 strategies that are up and running. It feels like a Niagara Falls of quarterly reporting.” To manage this data overload, they aim to develop a tool to analyze different strategies effectively.

Kearns assures us that the students are close to creating a solution that meets their reporting needs, ensuring compliance while safeguarding sensitive client data.

Aligning Goals with Technology

Himanshu Joshi, Senior Project Manager of Applied AI Programs at the Vector Institute, emphasizes the need for businesses to align their goals with available AI technologies. He notes the importance of determining which problems AI can address effectively.

Moreover, Joshi underscores the necessity of developing AI innovations within a responsible and ethical framework. Such frameworks ensure sensitive information is safeguarded while adhering to regulations.

To support this initiative, Vector offers seminars and boot camps aimed at making businesses "AI-ready." Recognizing that many small family offices lack tech expertise, Vector also places interns—often PhD or master’s students—within these organizations.

“In some cases, our collaborations result in new products that can be marketed to other companies,” Joshi mentions. “Remarkably, 60 percent of machine learning associates are hired as permanent employees.”

The Inflection Point for Family Offices

Craig Stewart, Executive Director of Applied AI Programs at Vector, asserts that family offices are at a critical juncture. They cannot afford to be indifferent regarding AI’s potential to transform their operations.

“Learn as much as you can and ensure that AI is part of your senior management and board discussions,” advises Stewart. “Family offices must stay informed on this technology. If they fail to explore their options, they risk falling behind in both domestic and international markets.”

Early Adoption: Northland Wealth Management

Arthur Salzer, CEO and CIO of Northland Wealth Management, illustrates the benefits of early adoption. His office was among the first in Canada to utilize Addepar, wealth management software launched in 2009. This software specializes in data aggregation and analytics, enhancing portfolio reporting.

However, Salzer cautions that AI isn't an instant solution for efficiency. He compares it to guiding interns through tasks. “It’s not as simple as flipping a switch,” he states. “You need to mold and direct the system for optimal use.”

Human oversight remains vital. For instance, understanding how machine learning software produces reports is crucial. “If discrepancies arise, we must know if it’s due to a data feed issue or incorrect inputs from our end,” Salzer explains.

Regulatory Framework

In early December, the Canadian Securities Administrators (CSA) published guidance on AI usage for registrants. This document emphasizes the importance of human oversight. The CSA highlights that current AI development stages do not support using AI exclusively for client investment decisions.

Managers of large investment funds have leveraged machine learning for a competitive edge, according to Sidney Shapiro, Assistant Professor of Business Analytics at Alberta's University of Lethbridge. The CSA's upcoming guidelines may ensure that advisors don’t rely solely on AI and consistently prioritize their clients' best interests.

The Limitations of AI in Investment Advice

Shapiro notes that generative AI applications, such as ChatGPT, can analyze data but may misuse qualitative insights for quantitative analysis. For example, he points out that asking ChatGPT how many times the letter "d" appears in "dividend" can yield incorrect results. While humans easily answer this, AI struggles due to a lack of contextual references.

“AI here isn't truly thinking; it's a massive recycling machine looking at past data,” he insists. “Past performance doesn’t dictate future results."

He likens relying on generative AI for financial advice to a doctor using it to diagnose patients: “Generative AI will offer broad advice but may not consider individual cases.” A skilled financial advisor combines reliable information with personal experience and the client’s unique needs—qualities that AI cannot replicate.

Original Article