AI models pass CFA exam’s toughest level, study finds
NEW YORK, UNITED STATES — A groundbreaking study reveals that artificial intelligence models have advanced to the point where they can now complete the notoriously difficult Level III of the Chartered Financial Analyst (CFA) exam.
The CFA designation, granted by the CFA Institute, is a globally recognized credential for financial professionals that requires passing three rigorous exams. The ethics, economics, and investment analysis exams have historically been a significant issue due to a long-term pass rate as low as 45%.
The possibility of artificial intelligence (AI) passing this high bar points to a seismic change, and some of the essential abilities that constitute financial expertise are about to be radically redefined. The charter is therefore a symbol of eminent accomplishment, and more than 200,000 charter holders worldwide have built careers based on this esteemed qualification.
The research, developed by experts from New York University’s Stern School of Business and GoodFin, demonstrates how AI can now navigate even complex essay questions, which were previously a major hurdle for machine learning systems.
AI passes CFA Level III mock exams
The research highlights AI’s ability to handle high-level financial analysis. For the first time, AI models have passed the CFA Level III mock exams, which are complex and focus on portfolio management and wealth planning.
This milestone was achieved by utilizing advanced large language models, such as o4-mini, Gemini 2.5 Pro, and Claude Opus, employing “chain-of-thought prompting” to answer both multiple-choice and essay questions successfully.
Previously, AI struggled with the CFA Level III exam, particularly with the essay component, which requires nuanced reasoning and interpretation of complex financial scenarios. This restriction was reported by research two years ago.
The recent advancement, however, represents a drastic transformation of AI technology, as these new models exhibit the capability of performing specialized, high-stakes analytical reasoning required to make professional financial decisions.
AI’s role in financial industry and human collaboration
While the research demonstrates AI’s growing proficiency in passing high-level exams, experts agree that AI is unlikely to replace human professionals in the financial sector.
Anna Joo Fee, Chief Executive Officer (CEO) and Founder of GoodFin, noted that, “I think there’s absolutely a future where this technology transforms the industry.”
Fee stressed the significance of human touch in decisions that require empathy and judgment, such as client-related and individualized financial advice. “There are things like context and intent that are hard for the machine to assess right now,” Fee said.
These aspects are essential for understanding non-verbal cues and human emotions, which AI cannot fully grasp.
“That’s where a human shines, in understanding your body language and cues,” she added.
As AI continues to increase its capabilities, the future of the financial sector may be a collaborative one, with AI handling all the computational work and humans bringing the emotional depth and subtle insights that technology cannot provide at present.
Future for AI and human professionals in finance
With the ever-evolving field of AI, there is an increasing presence of AI in the financial sector.
Fee highlighted how the human touch continues to play a crucial role in providing personalized services that AI models cannot do.
The study’s findings reveal that AI has a bright future in financial decision-making, though it will likely complement rather than replace human professionals in the field.
