In the so-called “Wild West” space of social media financial influencers (or “finfluencers”), FINRA regulators warned that investors in the self-directed market are particularly at risk of fraud and poor advice. 

During a panel at FINRA’s annual conference in Washington, D.C., this month, Morgan Stanley Wealth Management Workplace Compliance Director Megan Powers stressed that clients (and advisors) who use generative artificial intelligence tools like ChatGPT may be inadvertently influenced by self-fashioned finfluencers.

“(ChatGPT’s) learning from these social media influencers and blogs, et cetera. And that may or may not be relevant for an investor, right?” she said. “It’s about not only understanding what these social media influencers are saying, but where is the AI getting its information and what input is the investor putting into it?”

Powers argued that as AI wields greater influence among influencers (and in advice more broadly), a financial advisor should act as an intermediary for clients, noting aberrant behaviors and escalating concerns when needed. However, in the self-directed space, the oversight layer that an advisor provides may be nonexistent. 

Related:Stifel Reaches Settlement in Structured Notes Case

“Even with influencers in the social media space that have subscriptions or a big following, they say ‘this should not be considered investment advice,’ but the retail investor is strong, and we saw that back in 2021 with GameStop,” she said. “It’s harder for people to separate who has the background and experience versus someone who’s a novice that’s just looking for followers.”

The panelists also delved deeper into an April 2026 report from FINRA showing social media users and finfluencer followers were more likely to be victims of fraud despite conducting more due diligence when vetting financial professionals (among fraud targets, about 69% of social media users/finfluencer followers lost money, compared with about 29% of others).

In 2021, FINRA ran targeted exams on firms that recruited such finfluencers, and also released tips for broker/dealers working with influencers, including evaluating their backgrounds and prior social media activity for compliance and reputational risks and maintaining records of their public communications (many compliance experts found the tips borderline unworkable). 

In 2024, FINRA filed several enforcement actions resulting from the exam probing firms’ oversight of paid influencers.

During the panel, FINRA Vice President and Chief of Staff Sarah Green noted that finfluencers dispelling false or misguided information often aren’t acting out of malice, but during “their own investing journey” didn’t feel there were information sources from individuals with their background or life experience.

Related:FINRA to Review of ‘Higher-Risk’ Structured Products

“So they used that as an opportunity to start their own channel and share information,” she said, “There’s a whole swath of people who aren’t aware that they have stepped into a somewhat regulated space.”

According to Olivia Valdes, a senior principal researcher with FINRA’s Investor Education Foundation, AI posed a unique challenge to researchers gauging investor (and influencer) behavior. 

According to Valdes, investors often disbelieve information generated via AI. She recalled a study in which the foundation presented the same information to two groups of investors and told one group that it came from AI while the other was told it was prepared by a financial professional; the latter group trusted the information more.

“We see a lot of different sources acknowledge that there’s a trust issue with AI, but trust is not inhibiting use,” she said. “Of course, there’s a huge risk with AI, just like with social media and influencers; you don’t know what’s going on. If something goes wrong, what do you do?”

To Powers, investor education on the part of firms was all the more essential as agentic AI models continue to make waves, likely heading for consumer use soon enough. Powers stressed that firms should be telling clients what to look out for to prepare themselves.

Related:SEC Ends Decades-Old ‘Gag Rule’ in Enforcement Settlements

“We talked about the influencer today. The influencer tomorrow could be an AI agent that looks like a human, right?” she said. “So, that’s even riskier.”