The CFA Institute introduced a new framework to help investment professionals prepare for artificial intelligence’s impact on capital markets before changes occur.
The AI Transition Framework marks the first part of a broader research series examining AI’s expanding role in investment management. It highlights four forces already changing the industry: AI system capabilities, adoption depth among firms, shifting human-machine labor divisions, and broader market effects.
Four possible futures for AI in capital markets
The framework presents four scenarios for how markets might evolve as AI adoption grows.
In augmented markets, AI enhances efficiency in research, portfolio construction, and risk management without altering market structure. Human oversight continues, active management remains viable, and firms experience productivity gains alongside gradual fee reductions.
A competitive divergence scenario would create uneven adoption, with firms fully integrating AI gaining cost and scalability advantages. Those relying on older systems could face margin pressure, talent losses, and declining performance. Asset owners would likely categorize managers into AI adopters and non-adopters, while smaller firms might need to invest, partner, specialize, or leave the market.
Under platform convergence, AI capability spreads widely through shared infrastructure like foundation models and centralized data systems. This could reduce differentiation among managers, concentrate influence in a few dominant platforms, and diminish the informational edge that has historically driven returns.
The most transformative scenario, model-mediated markets, envisions AI systems handling signal generation, capital allocation, and risk calibration. Discretionary and systematic investing would merge, portfolio construction would become largely automated, and governance credibility—not analytical output—would become the primary differentiator.
Related: Court orders TD Bank repay $16M defrauded investors
A key concern is where firms adopt similar AI models, training data, and decision frameworks. This could reduce the diversity of market views that has historically maintained system resilience, creating what some call “monoculture risk.”
To address this, the framework identifies three governance priorities. First, maintaining auditable human oversight in hybrid decision systems. Second, developing validation methods for AI models that continue learning after deployment. Third, creating oversight mechanisms for shared analytical infrastructure as concentration risk shifts from firms to technology providers.
The framework belongs to a broader research series that will examine how the investment profession is evolving, how accountability functions in AI-mediated markets, and what safeguards are needed to preserve fairness and stability.
“The investment profession has always adapted within the capital market ecosystem, and this transition will follow the same pattern,” said Mona Naqvi, managing director of the CFA Institute Research and Policy Center and the paper’s author. “Understanding these structural changes early allows professional standards, governance, and market practices to evolve ahead of AI’s deeper integration.”
Naqvi noted that the implications go beyond efficiency. “AI’s integration affects price formation, capital allocation, and the financial system’s integrity and stability,” she explained. “As AI-driven analytical capability becomes more widespread, professional competence will depend increasingly on judgment, ethics, and responsible governance of complex systems.”
The shift in allocations reflects broader changes in investor behavior as sentiment improves.
Leave a Reply