AI tools hallucinate on finance-related queries up to 41% of the time. In March 2024, the SEC fined two investment advisory firms for lying about how they used AI. The regulatory environment has changed—and most wealth management sales teams have not caught up
.png)
Estimated Read Time: 4 minutes
A frequently cited benchmark in discussions of AI reliability in financial services: AI tools hallucinate on finance-related queries in up to 41% of cases. That number deserves to sit for a moment. In nearly half of finance-specific questions, an AI tool is generating output that is inaccurate, unsupported, or fabricated—and presenting it with the same confident tone it uses when it is correct. The client or advisor reading that output has no reliable way to distinguish a correct answer from a hallucinated one based on the AI’s presentation.
The types of errors are varied: wrong historical return figures, inaccurate descriptions of fund characteristics, fabricated regulatory requirements, incorrect tax treatment of investment products, and misrepresented fee structures. In a retail investment context, any of these errors has potential suitability implications. In a high-net-worth or institutional context, the stakes are proportionally higher. An advisor relying on AI-generated product summaries to prepare for a client meeting is preparing with a document that may contain multiple material inaccuracies.
The SEC’s March 2024 enforcement actions against two investment advisory firms marked a turning point. Both firms had represented to clients that AI was driving their investment decisions. Neither was telling the truth about how their AI was actually deployed. The combined $400,000 in fines was not the most significant outcome—the explicit framing of “AI-washing” as a securities law violation was. The Commission made clear that misrepresenting how AI is used in investment management is the same category of violation as any other material misrepresentation to clients.
The regulatory trajectory since then has been consistent. Securities class actions targeting AI misrepresentations doubled between 2023 and 2024. The SEC’s Division of Examinations incorporated AI oversight into its examination priorities. For registered investment advisers, the compliance implication is that AI tools used in client-facing or investment advisory contexts require the same level of documentation, oversight, and accuracy verification as any other compliance-sensitive process. The client interaction records that wealth management teams maintain are not just CRM data—they are, in a regulated context, part of the compliance infrastructure.
Beyond the regulatory exposure, there is a fundamental trust question. Wealth management is a relationship business built on the advisor’s credibility as a guide through complex financial decisions. An advisor who presents AI-generated product information as their own analysis and is later shown to be inaccurate has damaged something that is very hard to repair. The client’s reasonable expectation is that the advisor knows what they are talking about—and that expectation does not accommodate “the AI told me so” as an explanation.
High-performing wealth management advisors are not avoiding technology—they are using it in the right places. They automate scheduling, document preparation, and administrative workflow. They do not automate client communication accuracy or let AI-generated content reach clients without review. And they treat the records of every client conversation as an asset to be protected with the same care they apply to the portfolios they manage, because the relationship is worth at least as much as the AUM it represents.
