Does AI have behavioral bias?

From virtual assistants and voice recognition, to self-driving cars and ridesharing technologies, artificial intelligence (AI) is a unique and effectual method of completing everyday tasks. Behavioural biases is the saviour for humans in all spheres. This exclusive trait makes us indispensable from processes and tasks that could be automated to optimum efficiency. We might take a second more to spot opportunities or to pick a choose, but empirical data has proved our proficiency at tasks vis-a-vis coded softwares.

It was alarming to witness that an AI powered Equity ETF in the US had outperformed 87% of other comparable active managers, which means that only 13% of the managers actually used the exclusive trait over the benefits that a software entails. Though behavioural biases and personal gut might be the last strings to our existence in the financial markets, but even theory proves that biases always cause our decisions to be sub-optimal than rational.

This was essentially what the outperformance of AI powered ETF (Ticker: AIEQ) signalled. It grew by exquisite stock selection skills and not by timing the market to buy cheap. Selecting stocks entails significant subjectivity unanswered by crunching and juggling data points.

So have the AI programs been coded well enough to incorporate behavioural biases as well?