Mediaphotos | Getty Images
For people to pass the distinguished, three-part chartered monetary analyst exam, it sometimes takes round 1,000 hours of learning over the course of a number of years. New analysis discovered that the expertise underpinning a slew of synthetic intelligence fashions is now superior sufficient to pass even probably the most tough – Level III – mock exams in a matter of minutes.
The new study – developed by researchers from New York University Stern School of Business and Goodfin, an AI-powered wealth-management platform – evaluated 23 massive language fashions on their skill to reply a number of selection and essay questions on mock CFA Level III exams. They discovered frontier reasoning fashions, together with o4-mini, Gemini 2.5 Pro, and Claude Opus, have been in a position to make use of “chain-of-thought prompting” to efficiently pass.
Previous analysis, notably from two years in the past, had discovered AI may clear Levels I and II of the exam, however it struggled with Level III, as a result of essay questions. However, the expertise has developed so quickly that the researchers wished to know whether or not the fashions may deal with, “specialized, high-stakes analytical reasoning required for professional financial decision-making.” The third CFA exam is primarily targeted on portfolio administration and wealth planning.
“I think there’s absolutely a future where this technology transforms the industry,” stated Anna Joo Fee, founder and CEO of Goodfin, which contributed to – however didn’t pay for – the analysis.
Still, Fee stated she doesn’t suppose the AI will finally exchange the CFA.
“There are things like context and intent that are hard for the machine to assess right now,” Fee stated. “That’s where a human shines, in understanding your body language and cues.”