Certain fund managers have begun directing capital toward exchange-traded products powered by AI-driven stock-selection tools, on the basis that those systems can absorb and interpret market data at a speed and scale beyond what human analysts typically achieve — a position that sits at odds with the wariness, and in some cases outright resistance, that automated investment technology continues to draw from wider audiences.

At present, a handful of ETFs — numbering fewer than six in total — draw on automated, algorithm-driven systems to handle the full investment process, from gathering market intelligence through to picking individual stocks, yet the combined assets under their management amount to roughly $30 million against an industry-wide pool approaching $17 trillion. Adopted across the bulk of asset management firms to at least some degree, AI tools have been working their way into portfolio and investment decision-making — a shift that remains largely beneath the radar for most end investors, even as the technology's footprint in the sector grows from a narrow base.

A Mercer poll conducted earlier this year found that 91 per cent of asset managers questioned were planning to expand their reliance on AI-driven portfolio tools, while those same respondents noted they were still working through which responsibilities could be delegated to automated systems without risk.

A small cluster of funds built entirely around algorithmic decision-making have posted returns that outpaced broader market benchmarks — though their track records, in most cases, span only a matter of months. Those who manage such funds argue that performance figures will attract more capital to the approach over time. Tal Schwartz — who founded Ai Funds and serves as its chief executive — framed the investment landscape as one where participants had long faced a binary choice between passive index exposure and stock-picking by human hands, and noted that a third avenue, powered entirely by machine intelligence rather than human judgment, was the gap he set out to fill when the High Conviction US Equity AI-Managed ETF launched in August, a vehicle whose core is a proprietary machine-learning model built on market data spanning multiple decades.

ETF.com data put the fund's total assets at under $20 million at that point, though Schwartz noted the product had not yet had much time to build traction. Schwartz noted that AI processes larger volumes of data at greater speed than any human analyst while removing emotional and cognitive influences from investment decisions, adding that sceptics would likely come around only through demonstrated returns over time — a thesis his fund has so far supported, with a 2.7 per cent gain recorded since its 21 August launch against a 1.1 per cent rise posted by the S&P 500 over the same period.

Both products that FinQ — an asset manager based in Israel — brought to market in February have posted returns above the S&P 500's 11.0 per cent advance over the same stretch, clearing that index benchmark. Since its launch, FinQ's dollar-neutral US Large-Cap AI-managed ETF has posted a gain of 25.35 per cent, while the unhedged variant of the same fund stands 23.98 per cent above its opening price.

The Pacer S&P 500 3AI Top 100 ETF and the Pacer S&P World 3AI Top 100 ETF have posted gains of 10.34 per cent and 10.33 per cent respectively from their early-May debut dates, outpacing the S&P 500, which returned 5.8 per cent across the same window.

Investor appetite for these funds may be held back by their limited track records and above-market fee structures, practitioners in the field say. Jeff Judge, a financial planner at Chesapeake Financial Planners in Forest Hill, Maryland, reports that client demand for AI-driven products has not materialised, and draws a line between AI that supports human decision-making and AI that replaces it entirely. Todd Sohn, an ETF analyst at Strategas, argues that how any such strategy weathers a sustained market downturn will be the true measure of its durability.

Quantitative analysis, stock screening, and risk tracking are among the functions that asset managers at competing firms have built into a wave of ETF products marketed under artificial-intelligence branding. That measured stance is widespread across the industry, according to practitioners in the field — among them David Wright, who oversees quantitative investments at Pictet Asset Management in Geneva and noted that the firm had been deploying AI to train models for roughly seven or eight years and directing live capital through those systems for around four.

A fund under the Pictet AI Enhanced US Equity ETF name was brought to market by the firm that February. Drawing in $163 million worth of investor assets, the fund has delivered a 14.8 per cent return over that span — outpacing the S&P 500, which recorded 11.4 per cent across the same period.

To explain how artificial intelligence fits into his investment workflow, Wright reaches for the analogy of a commercial aircraft's autopilot system. Wright noted a personal reluctance to board any aircraft operating without a pilot in the cockpit — a reservation he said was echoed across much of the asset management industry, where firms are treading carefully as they incorporate AI into their operations. Dan Taylor, deputy chief investment officer of systematic at Man Group, told reporters that headlines about AI threatening civilisation did not keep him awake at night, though he acknowledged a genuine concern about whether the technology can reliably navigate the rarest, most unpredictable market conditions — the moments, he said, where outcomes at both ends of the spectrum tend to be decided.