For years, Delphia Technologies Inc. has claimed artificial intelligence drives its investment decisions for its tens of thousands of clients. Those claims, the U.S. Securities and Exchange Commission now says, were lies.
The SEC said Monday that Toronto-based Delphia has agreed to pay a US$225,000 fine in order to settle charges of making false and misleading statements about its purported use of AI. From 2019 through 2023, according to the SEC order against Delphia, the company claimed to put “collective data to work to make our artificial intelligence smarter so it can predict which companies and trends are about to make it big and invest in them before everyone else.”
But the commission says this never happened. “Delphia’s plan to utilize client data in artificial intelligence and machine learning never materialized,” the order says. “While Delphia did collect certain client data intermittently between 2019 and 2023, it never used that data with artificial intelligence or machine learning or otherwise used that data in any way as inputs into its investing algorithms.”
As investment advisers and portfolio managers promote AI as the latest and greatest means of maximizing returns, the case highlights a growing trend that the SEC refers to as “AI-washing” – the practice of using AI solely as a marketing tool.
Nicola Lacetera, a professor at the University of Toronto’s Rotman School of Management who studies digital technologies and the ethics and moral limits of markets, said AI is hard to define, which makes it easier for companies to mislead the public about their use of the technology.
“It can be easy to say you are using AI, to attract attention. Then people cannot figure out whether it is true or false,” he said. “It is part of the nature of this new technology as a whole, which is the question of how transparent or not it is.”
Investors appear to want more transparency. In an Ipsos survey conducted in November, 2023, on behalf of the Responsible Investment Association, nearly three in four individual Canadian investors – 74 per cent of respondents – said companies should be encouraged to “provide transparency and data on how they are using and investing in AI.”
As of Sept. 1, Delphia had about US$187-million in assets under management. Most of that total – US$180-million – was in five pooled investment funds. The remaining US$7-million was in roughly 29,000 individual retail accounts, managed through the company’s algorithm-based robo-advisory service.
In accepting the settlement, Delphia neither admitted nor denied the SEC’s findings. But, in response to questions from the SEC’s Division of Examinations in July, 2021, the company “admitted … that it had not used any of its clients’ data and had not created an algorithm to use client data,” the order says.
After promising to improve the accuracy of its claims and to remove any false statements from its website and marketing materials, the SEC says, Delphia “continued to make false and misleading statements” until as recently as August, 2023.
On the home page of the Delphia website, viewed by The Globe and Mail on Monday afternoon, the company says its “algorithm looks to spot trends in the data – and aims to get smarter the more transactions we see.”
“By training an AI on the data we buy, we attempt to spot trends before others catch on,” the website adds.
Delphia did not respond to multiple requests for comment.
SEC chair Gary Gensler said in a video posted to YouTube on Monday that AI is only the latest technology to be exploited for marketing. “When new technologies come along, we’ve also seen time and again false claims to investors by those purporting to use those new technologies,” he said.
“Everyone may be talking about AI, but when it comes to investment advisers, broker dealers and public companies, they should make sure that what they say to investors is true.”
Prof. Lacetera said investors should be wary of investment companies making broad, sweeping claims.
“Investment companies, any of them that are claiming to be better because they are smarter or use this sort of technology, people should take those claims with a grain of salt,” he said. “Just saying you use AI is too little. It is too generic, but it is very flashy.”
Jameson Berkow