As many as 91% of asset managers are using or plan to use artificial intelligence in their investment strategy or asset class research, according to the report produced by Mercer Investments.

But this use has been limited to relatively simple tasks like data analysis or idea generation. A much smaller number are relying on AI for more complex processes such as portoflio construction or rebalancing.

The report echoes a trend found elsewhere in the financial services world when it comes to the use of AI. To date, the majority of use cases have centred on labour-saving processes such as data analysis, rather than decision-making.

However, Mercer’s research suggests this may be changing. It found that 21% of the surveyed firms are planning to launch AI-driven investment strategies in the next 12 months.

According to Jo Holden, global head of investments research at Mercer, the step from applying AI to drive efficiencies to the implementation of more complex aspects of investment management will be “key to AI becoming truly transformative”.

The study involved interviews with 150 investment managers.