Category: Canada

As Canada moves toward stronger AI regulation with the proposed Artificial Intelligence and Data Act (AIDA), its southern neighbour appears to be taking the opposite approach.

AIDA, part of Bill C-27, aims to establish a regulatory framework to improve AI transparency, accountability and oversight in Canada, although some experts have argued it doesn’t go far enough.

Meanwhile, United States President Donald Trump’s is pushing for AI deregulation. In January, Trump signed an executive order aimed at eliminating any perceived regulatory barriers to “American AI innovation.” The executive order replaced former president Joe Biden’s prior executive order on AI.

Read more:
How the US threw out any concerns about AI safety within days of Donald Trump coming to office

Notably, the U.S. was also one of two countries — along with the U.K. — that didn’t sign a global declaration in February to ensure AI is “open, inclusive, transparent, ethical, safe, secure and trustworthy.”

Eliminating AI safeguards leaves financial institutions vulnerable. This vulnerability can increase uncertainty and, in a worst-case scenario, increase the risk of systemic collapse.

Read more:
The Paris summit marks a tipping point on AI’s safety and sustainability

The power of AI in financial markets

AI’s potential in financial markets is undeniable. It can improve operational efficiency, perform real-time risk assessments, generate higher income and forecast predictive economic change.

My research has found that AI-driven machine learning models not only outperform conventional approaches in identifying financial statement fraud, but also in detecting abnormalities quickly and effectively. In other words, AI can catch signs of financial mismanagement before they spiral into a disaster.

In another study, my co-researcher and I found that AI models like artificial neural networks and classification and regression trees can predict financial distress with remarkable accuracy.

Artificial neural networks are brain-inspired algorithms. Similar to how our brain sends messages through neurons to perform actions, these neural networks process information through layers of interconnected “artificial neurons,” learning patterns from data to make predictions.

Similarly, classification and regression trees are decision-making models that divide data into branches based on important features to identify outcomes.

Our artificial neural networks models predicted financial distress among Toronto Stock Exchange-listed companies with a staggering 98 per cent accuracy. This suggests suggests AI’s immense potential in providing early warning signals that could help avert financial downturns before they start.

However, while AI can simplify manual processes and lower financial risks, it can also introduce vulnerabilities that, if left unchecked, could pose significant threats to economic stability.

The risks of deregulation

Trump’s push for deregulation could result in Wall Street and other major financial institutions gaining significant power over AI-driven decision-making tools with little to no oversight.

When profit-driven AI models operate without the appropriate ethical boundaries, the consequences could be severe. Unchecked algorithms, especially in credit evaluation and trading, could worsen economic inequality and generate systematic financial risks that traditional regulatory frameworks cannot detect.

Algorithms trained on biased or incomplete data may reinforce discriminatory lending practices. In lending, for instance, biased AI algorithms can deny loans to marginalized groups, widening wealth and inequality gaps.

In addition, AI-powered trading bots, which are capable of executing rapid transactions, could trigger flash crashes in seconds, disrupting financial markets before regulators have time to respond. The flash crash of 2010 is a prime example where high-frequency trading algorithms aggressively reacted to market signals causing the Dow Jones Industrial Average to drop by 998.5 points in a matter of minutes.

Furthermore, unregulated AI-driven risk models might overlook economic warning signals, resulting in substantial errors in monetary control and fiscal policy.

Striking a balance between innovation and safety depends on the ability for regulators and policymakers to reduce AI hazards. While considering financial crisis of 2008, many risk models — earlier forms of AI — were wrong to anticipate a national housing market crash, which led regulators and financial institutions astray and exacerbated the crisis.

A blueprint for financial stability

My research underscores the importance of integrating machine learning methods within strong regulatory systems to improve financial oversight, fraud detection and prevention.

Durable and reasonable regulatory frameworks are required to turn AI from a potential disruptor into a stabilizing force. By implementing policies that prioritize transparency and accountability, policymakers can maximize the advantages of AI while lowering the risks associated with it.

A federally regulated AI oversight body in the U.S. could serve as an arbitrator, just like Canada’s Digital Charter Implementation Act of 2022 proposes the establishment of an AI and Data Commissioner. Operating with checks and balances inherent to democratic structures would ensure fairness in financial algorithms and stop biased lending policies and concealed market manipulation.

Financial institutions would be required to open the “black box” of AI-driven alternatives by mandating transparency through explainable AI standards — guidelines that are aimed at making AI systems’ outputs more understandable and transparent to humans.

Machine learning’s predictive capabilities could help regulators identify financial crises in real-time using early warning signs — similar to the model developed by my co-researcher and me in our study.

However, this vision doesn’t end at national borders. Globally, the International Monetary Fund and the Financial Stability Board could establish AI ethical standards to curb cross-border financial misconduct.

Crisis prevention or catalyst?

Will AI still be the key to foresee and stop the next economic crisis, or will the lack of regulatory oversight cause a financial disaster? As financial institutions continue adopt AI-driven models, the absence of strong regulatory guardrails raises pressing concerns.

Without proper safeguards in place, AI is not just a tool for economic prediction — it could become an unpredictable force capable of accelerating the next financial crisis.

The stakes are high. Policymakers must act swiftly to regulate the increasing impact of AI before deregulation opens the path for an economic disaster.

Without decisive action, the rapid adoption of AI in finance could outpace regulatory efforts, leaving economies vulnerable to unforeseen risks and potentially setting the stage for another global financial crisis.

Economic uncertainty stirred up by whipsaw U.S. tariff threats prompted BRP Inc. to push back its financial forecast for the coming year, with the trade limbo exacerbating weak consumer demand that drove the Ski-Doo maker to an earnings loss last quarter.

“There’s still the uncertainty around what’s going to happen on April 2nd, and I think that is influencing consumer behaviour,” said chief financial officer Sébastien Martel, referring to U.S. President Donald Trump’s pledge to impose 25 per cent tariffs on trade partners next week.

The U.S. has already hit Canada and Mexico with 25 per cent levies on goods that are not compliant with the North American free trade pact. The reprieve Trump granted on March 6 for items that do comply — a climbdown from blanket tariffs rolled out two days earlier — is also set to expire in one week.

Canada has hit back with its own duties on about $60 billion worth of American goods, and threatened tariffs on billions more if the U.S. does not back down.

“It’s difficult to call. It’s been choppy, and obviously with the uncertainty created by all of this, the consumers are holding back,” Martel told analysts on a conference call Wednesday.

“That uncertainty is a bigger overhang than the potential opportunity of buying a product with no tariffs today. It says a lot about the how the consumer is feeling.”

BRP swung to a loss of $44.5 million in the fourth quarter, down from a $302.8-million profit a year earlier.

As consumers and dealers bought less, North American retail sales at BRP dropped 21 per cent year-over-year in the quarter ended Jan. 31, largely due to lower demand for snowmobiles and market share loss in off-road vehicles. Its three-wheeled motorcycles saw retail sales fall about 30 per cent.

A continental trade war bodes ill for a company with factories in all three countries — the direct result of free trade agreements dating back decades. Some 60 per cent of BRP’s revenue stems from the U.S. Most of the inventory sold there is made in Mexico — 70 per cent of total production happens south of the Rio Grande — or Canada, where Ski-Doos and some of its Can-Am three-wheelers roll off the line.

“It could have a sizable impact if tariffs were imposed on all goods crossing the border,” said Martel.

Nonetheless, BRP beat earnings expectations in its latest quarter, triggering an eight per cent jump in its share price to $54.77 in late-morning trading on the Toronto Stock Exchange.

On a normalized basis, the company’s diluted earnings hit 98 cents per share in the fourth quarter versus $2.78 per share the year before. Analysts had expected 88 cents per share, according to financial markets firm LSEG Data & Analytics.

“All things considered, could have been much worse,” said Desjardins analyst Benoit Poirier in a note to investors.

Revenue for the three-month period fell 20 per cent to $2.1 billion from $2.6 billion in the same period a year earlier.

This report by The Canadian Press was first published March 26, 2025.

Companies in this story: (TSX:DOO)

Christopher Reynolds, The Canadian Press

As the U.S. trade war heat ups and increases pressure on an already battered loonie, some Canadians are wondering if this is a good time to exchange large amounts of Canadian dollars into greenbacks, or vice versa.

You may have some U.S. dollars and want to convert them to Canadian currency at today’s favourable rate. Others may be diversifying part of their portfolios to be denominated in U.S. currency as part of their investment strategy.

Whatever the reason, here’s how to exchange large amounts as efficiently as possible.

When you look up an exchange rate, the first quote you’ll likely see is known as the spot rate. This is the rate that big institutions like banks get when they trade large sums of cash with each other.

However, that’s not the rate that you get when you walk into a bank branch. Every currency has two prices: a buy price and a sell price. This is known as the “forex spread” of that currency. It’s also how your bank makes money – they buy currency at a lower price than the spot rate, then resell it at a higher one.

Let’s see how this works in practice. At the time of this writing, the CAD/USD spot rate is 69.85 US cents per Canadian dollar. I walked into a bank branch and tried to purchase U.S. dollars. The price they quoted was only 68.16 US cents. This represents a 2.5-per-cent forex spread.

Now, this might not be a big deal if you’re only exchanging $100. That spread only costs you $2.50. But if you’re exchanging $1,000, then the cost rises to $25. The more you exchange, the more you pay.

Fortunately, there’s a strategy to exchange large sums of cash without this spread using your trading account. It’s called Norbert’s Gambit. Here’s how it works.

There’s an ETF with the symbol DLR. This is the Global X US Dollar Currency ETF, which is a fund that holds cash in USD. What’s special about DLR is that it’s listed on the Toronto Stock Exchange in two different currencies: DLR (CAD) and DLR.U (USD).

Because the free market, rather than your bank, determines the price of these two securities, DLR and DLR.U trade at the actual exchange rate without a forex spread. So, these funds can be used to exchange any amount of CAD and USD at that better rate.

Let’s say you had $10,000 of CAD that you wanted to exchange for USD. At the spot rate you would get US$6,985, but the bank only gives you US$6,816, meaning you lose US$169 to the forex spread.

Now let’s see what happens with Norbert’s Gambit. As of this writing, DLR is trading for $14.62 per share in Canadian currency. At that price, $10,000 buys you 684 shares.

Now you contact your bank or brokerage and submit a “journaling” request, which transfers your shares of DLR to the USD version, DLR.U. Some banks allow you to submit this request online, others require you to call into customer support. Journaling can take two to four business days to complete.

After your journaling completes, you now have 684 shares of DLR.U sitting in your trading account. At DLR.U’s current price of US$10.21, selling these shares would net you US$6,983.64. This is now only $1.36 off from the spot price, and represents a forex spread of only 0.02 per cent, which is nothing compared with the 2.33 per cent the bank charges.

A few caveats you should know before doing this.

First, it requires your bank or brokerage to support journaling shares. Most, but not all, brokerages support this.

Second, it requires your bank or brokerage to support multicurrency trading accounts. The Big Six Canadian banks all allow this, though you may need to contact customer service to enable it.

Third, there may be a small fee to buy or sell the shares, or to process the journaling request. Questrade, for example, will start charging $9.95 per journaling request starting in April. However, these fees are flat and so don’t change depending on how much you convert. That’s why it makes more sense to use this strategy to convert larger sums.

And finally, journaling takes three to four business days to complete. Your exchange rate is determined when you buy or sell the Canadian-denominated DLR, so if you’re converting CAD to USD, the conversion rate is set immediately when you buy DLR. However, if you’re converting USD to CAD, the rate is determined after the journal completes, so the exchange rate may fluctuate during that time – leaving yourself exposed to the downside if CAD rises during those few days. (On the plus side, you could end up with an even better deal if the greenback goes up.) If you need the money right away, you may have to just suck it up and pay the forex spread.

Even with all these caveats, if you find yourself exchanging large amounts of CAD for USD, learning how to use Norbert’s Gambit could save you hundreds or thousands of dollars in foreign exchange fees over time. And any money saved in fees is money that can be put to better use elsewhere.

Kristy Shen and Bryce Leung retired in their 30s and are authors of the bestselling book Quit Like a Millionaire.

Canada-based biotech firm Willow Biosciences (TSX: WLLW) (OTCQB: CANSF) this week reported a net loss of C$6.1 million despite posting a roughly threefold increase in revenue in 2024. The company also appears poised to close down entirely, with operations slated to cease as soon as the sale of a subsidiary goes through.

Willow said in a press release that CEO Dr. Chris Savile and Senior Vice President of Research and Development Dr. Trish Choudhary will be terminated as soon as the company’s sale of Epimeron USA  closes, “as the Company will no longer be carrying on active operations.”

Willow announced the sale of the subsidiary to an unnamed buyer in the United Kingdom for US$3.3 million. That’s expected to improve Willow’s cash on hand from C$333,000 at the end of 2024 to C$1.1 million, the company forecast, while most of the proceeds from the sale will be used to pay off debts.

Even the sale, should it close, won’t be enough to keep Willow listed on the Toronto Stock Exchange, the company said in a press release, and shared that it expects to be suspended from the exchange.

The sale is expected to close by the end of April.

According to the company’s quarterly financial filing, Willow’s revenues increased substantially year-over-year to C$4.6 million from C$1.1 million, and it cut its annual losses by more than half, to C$6.1 million from C$13 million.

As of Dec. 31, Willow had C$2.4 million in total assets, including C$333,000 in cash, against C$2.9 million in total liabilities and a whopping C$127.8 million deficit.