The terminology used by investment firms to explain their technologies has consistently lagged behind its real capabilities. The discrepancy between marketing and reality carries a particular kind of weight in investment management, because the promises are directly linked to decisions people make with their savings.
This is not exclusive to finance; it also occurs in pharmaceuticals, energy, and consumer products. The Securities and Exchange Commission made the decision to begin tracking that disparity in March 2024. It discovered a gap large enough to qualify as fraud in at least two instances.
| Category | Details |
|---|---|
| Enforcement Action | SEC AI Washing crackdown — March 2024, first enforcement of its kind |
| Firm One | Delphia (USA) Inc. — fined $225,000 |
| Delphia’s Claim | Used collective user data to train AI that could “predict which companies and trends are about to make it big” |
| Firm Two | Global Predictions Inc. — fined $175,000 |
| Global Predictions’ Claim | Advertised “expert AI-driven forecasts” on its website and social media channels |
| Total Fines | $400,000 combined |
| Violation | SEC Marketing Rule — false and misleading statements about AI capabilities |
| Regulator Term | “AI washing” — making inflated or false claims about artificial intelligence use |
| Broader Context | Part of intensified SEC scrutiny of AI-related marketing in investment industry |
| Further Reading | Analysis at Financial Times AI coverage |
Investors and the general public were informed by Delphia (USA) Inc. that the company employed user-contributed data to train artificial intelligence that could identify emerging trends and firms. The pitch was persuasive in the same manner that current AI pitches are persuasive: it was based on actual ideas, using believable language, and described a result that enough people want to think is possible.
The AI performing this task either didn’t exist in the form reported or couldn’t accomplish what the company promised, according to the SEC’s inquiry. Delphia had to stop making the claims and pay a $225,000 punishment. As is customary in SEC settlements, the company never acknowledged nor refuted the findings, which provides no insight into internal culture or intent.
A comparable operation was conducted from a different perspective by Global Predictions Inc. Its website and social media platforms promoted what it called “expert AI-driven forecasts”—a term that sounds specific enough to have meaning but is ambiguous enough that proving or refuting it necessitates precisely the kind of regulatory scrutiny Global Predictions apparently hadn’t expected to receive.
The SEC imposed a $175,000 punishment after determining that the representations were deceptive and applying the Marketing Rule. When the Delphia penalty was added, the total amounted to $400,000. This amount would not bankrupt either company, but it is significant enough to indicate that the enforcement action is genuine rather than fictitious.
The phrase “AI washing” was taken by the SEC from the environmental parallel of “greenwashing,” which is the practice of disguising routine corporate operations in terms of sustainability in order to appeal to a specific group of investors. The analogy is fairly appropriate.
Both entail attaching a high-value concept to a product that doesn’t fully embody it, take advantage of the fact that most investors lack the technical expertise to independently verify the claims, and create markets where operators who are truly capable are at a disadvantage because they can’t compete with rivals who are willing to exaggerate their offerings.
The distinction is that in the context of greenwashing, where you can at least quantify emissions, AI’s opacity makes verification more difficult than nearly anything.
It is worthwhile to consider the motivations behind the behavior. For years, the investment management sector has witnessed increasingly large returns from quantitative and algorithmic firms. The narrative surrounding AI, which claims that it can identify patterns in data that human analysts miss and process information at speeds and scales that no team can match, has put significant pressure on traditional managers to claim involvement in that narrative regardless of whether they have actually built the infrastructure to support it.
A company that discloses to potential customers that it use AI receives different attention than one that doesn’t. The SEC is increasingly evaluating AI capability claims as a separate category of marketing risk rather than just another line in a disclosure form, but this pressure does not absolve the deception.
It’s difficult to ignore the fact that the first wave of AI washing enforcement targeted very small businesses. The bigger question is still mostly unanswered: what is being said about AI capabilities at companies that manage billions of dollars and have compliance departments that are sophisticated enough to write things carefully? The SEC’s actions in March 2024 were the first of their sort. They most likely won’t be the last, and further cases might involve figures that make $400,000 appear to be a rounding error.
