LONDON, 4 March 2026 – Artificial intelligence is rapidly transforming the investment industry, with algorithms increasingly capable of analysing vast amounts of financial data, predicting market trends and even replicating the trading behaviour of professional fund managers. Yet despite the growing power of machines, human investors may still retain an edge, provided they remain unpredictable in their decision-making.
A new analysis highlighted that AI systems are exceptionally good at solving mathematical and statistical problems, particularly those involving structured financial data. In areas where human investors often display behavioural biases, such as misjudging probabilities or falling for the “gambler’s fallacy”, algorithms frequently outperform people.
However, when investment decisions require judgment, ambiguity and interpretation, machines often struggle. This limitation may leave room for experienced fund managers who rely on qualitative insights, intuition and unconventional thinking.
AI Can Replicate Many Investment Decisions
Recent academic research demonstrates how quickly AI systems are catching up with human investors. One study led by Harvard Business School professor Lauren Cohen trained an artificial intelligence model using historical data from U.S. mutual fund managers.
The results were striking: the AI system was able to predict roughly 71% of the trades made by fund managers, effectively replicating the majority of their investment behaviour.
This finding suggests that many investment strategies, particularly those based on rules, patterns or widely available information, can be easily duplicated by machines. For asset managers relying heavily on systematic or formula-based approaches, this could erode their competitive advantage in an increasingly automated financial market.
The Alpha Comes From What AI Cannot Predict
Yet the same research also revealed a crucial insight. The bulk of investment outperformance, often referred to as “alpha”, did not come from the predictable trades that AI could replicate.
Instead, it came from the 29% of trades that the AI could not forecast, reflecting moments where fund managers exercised unique judgment or deviated from standard patterns.
In other words, the value of human investors lies precisely in the decisions that machines cannot anticipate. Fund managers who follow rigid rules or conventional strategies may become increasingly replaceable, while those who rely on original thinking and flexible decision-making may continue to outperform.
AI Also Inherits Human Bias
Another challenge for artificial intelligence is that it often learns from historical datasets generated by human behaviour. This means AI models can sometimes inherit the same biases and errors present in the data used to train them.
Research from Boston University found that AI systems perform rationally when mathematical solutions exist, but they can replicate human cognitive biases when decisions require interpretation or incomplete information.
As a result, AI may not always provide superior judgment in complex investment scenarios where market narratives, geopolitical developments and corporate strategies play a role.
The Future of Investing in the Age of AI
For global financial markets, the rise of artificial intelligence is likely to reshape the asset-management industry in the coming decade. Machines will increasingly handle data-driven analysis, portfolio optimisation and routine trading strategies, reducing costs and improving efficiency.
However, the research suggests that human investors may still maintain an advantage, at least for now, in situations requiring creativity, qualitative reasoning and unconventional thinking.
Ironically, this means the best defence against algorithmic competition may be something machines struggle to replicate: human unpredictability.
For investors navigating the age of AI, the lesson may be clear, the more formulaic the strategy, the easier it becomes for machines to copy it.
