New framework warns converging AI models may erode the diversity markets lean on for resilience
Artificial intelligence stands to reshape how capital is priced, allocated, and governed, not simply how fast investment firms process information.
The CFA Institute Research and Policy Center makes that argument in a new paper launching a research series on AI's integration into capital markets and the investment profession.
The publication, Artificial Intelligence and the Future of Finance: A Framework for Structural Change, introduces what the institute calls its AI Transition Framework, meant to help investment professionals, industry leaders, and regulators engage with AI-driven structural change rather than react to it after the fact.
The paper argues the shift reaches into price formation, capital allocation, and the integrity and stability of the financial system extending well beyond the productivity gains that dominate most industry discussion.
At the centre of the framework are four structural forces the paper says are already in motion: capability, adoption, substitution, and recomposition.
Capability expands what is technically possible, adoption determines how widely those tools are embedded, substitution reallocates decision-making between people and machines, and recomposition reflects the cumulative effect on market structure.
The institute frames these as interacting pressures rather than a fixed sequence.
From those forces, the paper sketches four possible future states for capital markets.
In augmented markets, AI improves workflow efficiency without materially reshaping market architecture.
Competitive divergence describes uneven adoption that widens gaps in performance, cost, and positioning across firms.
Platform convergence sees AI adopted broadly through shared infrastructure, compressing differentiation and concentrating influence within common analytical platforms.
In model-mediated markets, AI systems take primary responsibility for signal generation, allocation decisions, and risk calibration across large segments of capital.
For active managers, the paper suggests the character of skill could shift.
As the cost of generating structured analysis falls and analytical tools become widely available, differentiation may depend less on producing information and more on system design, data governance, model oversight, and institutional integration.
Mona Naqvi, managing director of the CFA Institute Research and Policy Center and the paper's author, said the profession has always evolved with the capital markets and this shift is no different.
She said the aim is to grasp the structural changes early so standards, governance and practice evolve ahead of AI rather than in response to it.
She added that the nature of professional competence itself is likely to change.
Naqvi said that as AI analysis becomes more embedded in investment decisions, professional competence will depend more on judgement, ethics and the responsible governance of complex systems.
The paper also introduces a concept it terms cognitive convergence, describing the growing alignment of AI models, data, and decision frameworks across institutions and the risk that correlated outputs compress the interpretive diversity markets rely on for resilience.
Model oversight, explainability, accountability, and institutional resilience, the institute said, will be central to maintaining trust as analytical authority becomes more model-mediated.
