AI Investment Advisors: So Many Risks, so Little Upside
Over the last year, and increasingly over the last months, adding the word "AI" to every product or service under the sun has become the ultimate marketing gimmick. There is even a term starting to emerge for it, “AI washing”, and it describes the growing trend of firms intentionally overstating or misrepresenting their technological edge in order to appear more innovative to investors.
The financial industry has not been immune of course, and there have even been legal cases already: In 2024, the SEC signaled a major crackdown on this practice, charging firms like Delphia Inc. and Global Predictions Inc. with making "false and misleading" claims about using AI to inform investment decisions when their systems were, in fact, far more primitive.
It doesn’t stop there, however. The AI investment trend has broken containment and is already gaining considerable traction on social media and other online platforms. From TikTok influencers promising "1,000% returns with my custom GPT" to sophisticated-looking YouTube channels showcasing "automated wealth bots”, promises of effortless, machine-driven, “guaranteed” profits are everywhere. Even more insidiously, some bad actors are using AI to generate deep-fake videos, or professional-looking white papers and totally made-up technical charts, to sell worthless tokens or to set up pump-and-dump trades. Deception is everywhere, which requires investors to pay close attention and question everything.
What AI can and cannot do (for now at least)
As our clients know well, at BFI, we are all for embracing new technology and we are not afraid of changing with the times. However, we do so carefully and only after thinking long and hard about all the potential risks and the benefits to the investors whose interests it is our duty to serve. AI is definitely one of the most exciting and likely game-changing new technologies of our generation. Even the most advanced models and tools out there today haven’t even begun to scratch the surface of what the (not-so-distant) future actually holds.
That being said, it is also becoming increasingly clear to us that the nascent industry is being surrounded by a certain level of hype, a fair amount of overpromising and underdelivering, and a lot of confusion, misconceptions and overall lack of understanding by the general public (which includes non-technically trained investors). When you don’t really understand how something works, it is hard to spot when it really doesn’t work.
When you follow an AI advisor blindly, you aren't just trusting a computer with your money, but you are also trusting the data it was fed and the ethics and motives of the person who prompted it. For instance, AI investment tools are often trained on historical snapshots. This means that in a volatile market where a single interest rate decision by the Fed or a surprising geopolitical event can shift prices in milliseconds, an AI relying on old data (even if it’s from yesterday) is effectively flying blind.
What’s more, a machine can recognize a pattern, but it cannot recognize a "Black Swan" event. This requires combining finance and investing knowledge with lessons from history and a deep understanding of human behavior, irrational choices, herd and bandwagon effects, and all the other mistakes humans keep making when panic or greed takes over.
A 2025 study by the University of St. Gallen found that AI is excellent at interpreting data, but it is very poor at the initial step of information retrieval. Two-thirds of all AI errors in financial analysis were caused by models failing to find the "needle in the haystack" within complex reports, as they struggle to find objective, deep-layer data, and then default to the most "available" information out there. They tend to take the path of least resistance by repeating what is most frequently mentioned online and as a result, they just end up parroting popular biases.
It is also important to note that unlike a human, a machine also doesn’t have a legal or moral obligation to protect your life savings. Or to play by the rules for that matter: Apollo Research, a UK AI safety organization, demonstrated in 2023 that an AI prompted to act as a trader for a fictitious financial investment company autonomously performed insider trading and subsequently lied to its own users to conceal its actions.
Human risks
Perhaps more dangerous than the risks stemming from AI limitations and “hallucinations” are those originating from the humans who control them. Investors really need to think “what’s in it for them” every time they see someone online touting their infallible AI investment tool.
To start with, it is important for investors (especially those not familiar with the dynamics of the social media economy) to understand that for a social media creator, the primary "product" is not a profitable portfolio. It is attention and clicks, views and comments, and how often their content is shared. TikTok and YouTube algorithms prioritize "high-arousal" content, which as one might expect, rewards outrageous or exaggerated claims over sound, boring financial advice. In other words, the creators in this case are using the technology to put out AI videos at a very low cost and with basically zero effort or research, with the sole goal of growing an audience and getting paid by ads on the platforms they are using.
Another motive might be affiliate marketing, with creators pushing their followers to sign up to a specific trading platform (usually unregulated), while the AI "advisor" is programmed to keep you trading frequently to generate commissions rather than making you money.
Finally, there’s the “black box” approach too, where the humans behind the AI advisor are disclosing next to nothing about how it actually works, but they are guaranteeing incredible profits if you just trust them with your funds. In late January, the FTC resolved a major case against Growth Cave, a company that allegedly used deceptive AI claims to lure investors into "automated" systems that were essentially empty shells.
Overall, investors need to be very cautious in this environment and as always, if something sounds too good to be true, it probably is. In a world where you can’t really trust your eyes anymore, you have no choice but to fall back on good old-fashioned common sense.
