Regulators are signaling that they’re watching financial institutions’ use of AI closely. MirrorWeb’s Harriet Christie explores what a recent report from FINRA means about the future of AI in financial services.
Both the SEC and the Financial Industry Regulatory Authority (FINRA) have shown they’re taking an interest in how businesses in the financial sector employ AI tools, and regulators’ skepticism is proving to be a counterweight to the industry’s increasing appetite for the technology.
Artificial intelligence
In its annual regulatory oversight report for 2024, FINRA classifies AI as an emerging risk, recommending firms consider its pervasive impact and the regulatory consequences of its deployment.
When you break down the ways in which marketers can leverage ChatGPT, for example, it becomes clear how effective the tool has become. Not only can it draft social media posts and website copy, it can also optimize them based on SEO, trending keywords or other relevant metrics. This saves marketers an incredible amount of work and will tempt stretched workforces in need of a lifeline.
Unfortunately, those teams might not be equipped to check the generated output thoroughly, which is especially problematic in the context of chatbot hallucinations, which is another way of saying AI can make things up. Without the correct checks and amendments, a brand’s tone of voice and clarity of messaging can be compromised. More worryingly, so can its factual legitimacy.
The SEC has already clarified that advisers themselves are responsible when issues arise after AI tools are used for investment recommendations, and it’s clear FINRA shares many of the same uncertainties. On a podcast dissecting the 2024 report, Ornella Bergeron, FINRA senior vice president of member supervision, said that despite the operational efficiencies afforded by developments in AI, there are worries.
“While these tools can present really promising opportunities, their development has raised concerns about things like accuracy, privacy, bias and intellectual property,” Bergeron said. “So far, firms are being very cautious and thoughtful when considering the use of AI tools, and before deploying new technologies. So while for this year’s report there was not a lot in the AI section by way of specific roles or observations, this is likely a topic we’ll be seeing a lot more about in the future.”
SEC’s Enforcement Head: It’s Time for ‘Proactive Compliance’
What Gurbir Grewal’s recent statements forecast for compliance leaders in 2024
Read moreDetailsPublic-facing communications
Two examination priorities for the SEC this year are AI and the new marketing rule, the transformative guidelines that came into effect in November 2022. These two priorities are intrinsically linked, with generative AI threatening to have a major impact on compliance with the marketing rule.
This rule redefined what constituted an advertisement to include a firm’s digital communications and stipulated that investment advisers may not circulate any advertisement that:
- Includes untrue statements and omissions
- Includes unsubstantiated material statements of fact
- Includes untrue or misleading implications or inferences
- Fails to provide fair and balanced treatment of material risks or material limitations
- Fails to present specific investment advice in a fair and balanced manner
- Cherry-picks performance results or otherwise presents performance in a manner that is not fair and balanced
- Is materially misleading
The first point is immediately compromised by AI’s propensity to mistake fiction for fact. ChatGPT, for example, is often more concerned with giving answers that sound coherent than ones that are accurate. And as for the second through seventh points, which use more subjective language, if AI tools can’t even tell an objective truth, can we really expect them to satisfy criteria set forth by regulatory bodies?
And don’t forget, as with the marketing rule, FINRA Rule 2210 encompasses electronic communications, meaning websites and social media channels will be held to the same standard as written brochures, TV advertisements and emails.
Harriet Christie is chief operating officer at MirrorWeb. She graduated from the University of Sheffield in 2010, with a B.A. in management accounting, entrepreneurship, business law, BSR, HR. She entered the tourism space, starting as an accounts executive at LateRooms.com, and earning the title of global accounts manager within three years. She occupied this role for a further five years as the business continued to evolve and flourish, before taking up her role as a key account manager with MirrorWeb, a data-archiving solution based in Manchester. Harriet was appointed operations director in 2020. Since then, she has helped oversee the evolution of the MirrorWeb product and service offering, as well as the business' impressive growth since her taking on the role. 
