Artificial intelligence has quickly moved from a novelty to a widely used tool in the investment world, with retail traders increasingly turning to chatbots for ideas, analysis, and even trade execution guidance.
While AI can streamline research and help process complex data, wealth advisors should be aware of the significant risks when clients use public AI models as a primary source of investment advice.
Recent trends show a growing number of individual investors—many from online trading communities like Reddit’s Wall Street Bets—sharing stories of how AI tools, including ChatGPT, have “helped them become better traders.” Some cite impressive results, like faster data analysis or more efficient screening of opportunities. For example, trader Erik Smolinski says AI has allowed him to replicate “six years of learning” in far less time by using it primarily for dissecting market data.
Yet, for every success story, there are far more cautionary tales. Advisors should be prepared to educate clients on where AI advice can go dangerously wrong.
The core issue: Public AI isn’t built for fiduciary-grade guidance
Markus Levin, cofounder of blockchain firm XYO, notes a fundamental limitation: public AI models often rely on a narrow, self-reinforcing set of publicly available sources when assessing emerging or niche investments. “For early-stage companies, these models tend to pull from company press releases, founder posts on X, Reddit threads, and managed media appearances,” he explains. This creates a risk of echo chamber analysis—where the AI recycles the same overly positive narratives without critical evaluation.
Levin stresses the difference between the sophisticated, proprietary AI models used by trading firms and the publicly available tools accessible to retail investors. “Retail traders are asking public LLMs questions like, ‘What project will 5–10x this year?’ and treating the answers as actionable insights. That’s where the real danger lies,” he warns.
When AI turns investing into “algorithm-assisted gambling”
Eric Croak, president of Croak Capital, is blunt about the dangers. “AI is dangerous with crypto and options specifically because it struggles to explain asymmetric risk in real terms,” he says. For example, these models may entirely omit key considerations like tax implications, liquidity constraints, or the probability distribution of returns. The result can be an overconfidence in trades that carry significant downside.
Croak’s description—“algorithm-assisted gambling”—captures a concern many advisors share: without proper context and risk disclosure, AI can encourage speculative behavior under the guise of data-driven insight.
No accountability means clients bear all the risk
Linda Ta Yonemoto, a personal finance educator, highlights another problem: accountability. “AI tools have zero obligation to a board, to regulators, or to the law. If AI advice leads to tax penalties or investment losses, you are entirely on your own,” she says. This lack of recourse means that even clearly flawed guidance—if followed—can have long-term consequences for a client’s portfolio and financial plan.
Recognizing red flags in AI-generated investment advice
For wealth advisors, it’s critical to be able to identify when AI-generated guidance is likely unreliable. Experts point to several warning signs:
-
Polished, confident language without specifics
Croak warns that the “biggest red flag” is advice that sounds overly clean and confident but lacks concrete details—such as specific numbers, timeframes, or cited sources. “If it reads like a marketing blog instead of an investment memo, that’s your warning,” he says. This kind of language often indicates the AI is generating persuasive-sounding content rather than rigorous analysis. -
One-size-fits-all recommendations
Tim Newell, founder and CEO of GreenFi, stresses that good advice—whether from a human or AI—should always start with the investor’s unique circumstances. “The biggest red flag is advice that’s too general,” he says. AI that doesn’t request or incorporate personal financial details is inherently limited and potentially misleading. -
Overconfidence without attribution
Jake Falcon, founder and CEO of Falcon Wealth Advisors, cautions against AI models that provide definitive predictions without noting uncertainty or data limitations. “Reliable advice should include context, limitations, and risk disclosures. Many AI outputs skip this entirely,” Falcon says. A lack of cited sources or supporting evidence should be a clear signal for skepticism.
The role for wealth advisors: AI as a tool, not a substitute
For RIAs and financial advisors, the takeaway is not to reject AI outright but to frame it as a supplemental resource rather than a primary decision-maker. Used well, AI can assist with:
-
Data aggregation and filtering
-
Preliminary market screening
-
Summarizing complex filings and reports
-
Generating alternative scenarios or stress tests
However, the final recommendations—especially for client portfolios—must be informed by a holistic view of their financial picture, risk tolerance, tax situation, and long-term goals. These are areas where public AI models have no built-in capacity for true personalization.
Educating clients on AI’s limitations
Advisors may want to establish clear talking points for clients experimenting with AI-driven advice:
-
Explain the data sources: Help clients understand that public AI models train on a mix of credible and less credible sources, and cannot always differentiate between the two without explicit human review.
-
Highlight the lack of fiduciary oversight: Make it clear that AI outputs are not subject to regulatory standards or fiduciary duty.
-
Stress-test the advice: Encourage clients to treat AI suggestions as hypotheses that require verification—through fundamental analysis, due diligence, and alignment with their plan.
-
Watch for emotional triggers: Many AI outputs can be persuasive in tone, creating a false sense of certainty. Teach clients to separate style from substance.
Balancing the conversation
There’s a risk that dismissing AI entirely could alienate clients who are curious about the technology. A better approach may be to integrate AI into the advisory process in a controlled, transparent way. For example, advisors might run AI-generated market screens and then explain how they confirm or challenge those findings with proprietary research. This both validates the client’s interest and reinforces the advisor’s role in risk management.
Looking ahead: Regulation and best practices
As AI becomes more embedded in personal finance, there will likely be growing pressure for regulation—particularly around transparency of sources, attribution of data, and the disclosure of model limitations. Until that happens, advisors remain the primary safeguard against poor decision-making.
In the meantime, best practices for integrating AI into investment workflows may include:
-
Using AI only for preliminary screening, not final allocation decisions.
-
Documenting when and how AI-generated insights are incorporated into recommendations.
-
Keeping a “human in the loop” for every critical investment call.
-
Continually testing AI output for accuracy and bias.
The bottom line for advisors
AI can be a powerful addition to the research process, but it is not a substitute for the depth of judgment and fiduciary responsibility that wealth advisors bring. The real danger lies in uncritically accepting AI’s outputs as fact—particularly when they are generalized, overconfident, or devoid of concrete, attributable data.
As more clients turn to AI for financial guidance, RIAs have an opportunity to position themselves as both interpreters and gatekeepers—leveraging the efficiencies AI offers while protecting portfolios from its blind spots. In this environment, the most valuable role of the advisor may be not just selecting the right investments, but ensuring the right process is used to choose them.
