Banks that fail to embrace AI risk losing up to $170 billion (£127.5bn) in profits, a new report from McKinsey has cautioned, warning that those institutions lagging in the era of AI will soon become irrelevant.

The consulting firm’s latest Global Banking Annual Review argues that agentic AI, in particular, has the potential to ‘radically reshape’ the banking sector as customers begin to integrate it into their finances, but any benefits for the banks themselves are unlikely to be felt across the industry.

While banks that move quickly on agentic and genAI could see equity returns jump by 4%, those slow to adapt risk long-term profitability decline by the same amount.

Without a strategic shift, McKinsey warns that banks could see global profit pools shrink by $170 billion, around 9%, over the next decade, potentially dragging average returns below the cost of capital. 

Driving this huge loss of profit that McKinsey heralds is growing consumer trust in AI to manage personal finances.

According to the report, $23 trillion (£17.2tn) of the world’s $70 trillion (£52.5tn) in consumer deposits currently sits in checking accounts offering near-zero interest, with the rest earning only marginally more.

McKinsey suggests that if just 5–10% of those balances were to shift toward top-of-market rates, prompted by AI-powered financial agents taking action on consumers’ behalf, banks could lose up to 20% of their deposit-based profits.

There is some good news for bankers, however, with McKinsey predicting that early adopters of AI are likely to see their costs for retail banking functions fall by at least 15%.

The report forecasts that AI will have the biggest impact on front office functions, with costs predicted to fall by 30% in the most likely scenario, followed by support functions (25%) and operations and risk management (25%).

Recommended reading

• Can AI Deliver the Next Big Leap in Banking?
• Most UK Banks Use AI, But Consumer Concerns Remain
• Report: Top Banks Advancing AI 2.3x Faster Than the Rest

However, these savings, though significant, will be undercut by competition among the banks themselves. A spike in technology costs, potentially rising by more than 70%, coupled with higher cybersecurity outlays and more fraud prevention spending, will see just a handful of pioneering banks reinvent their business models to defend against AI transformation.

The rest will see a decline in their market share and returns, with fierce competition quickly eroding meaningful gains. 

“In an era of falling revenues, banks sorely need productivity gains and could potentially get them from AI. But AI is a double-edged sword, likely to bring cost savings as well as disruption,” says the report.

“Early adopters will be able to secure a lasting advantage over slow movers. Given these are still the early days of agentic and genAI, it is imperative to use surgical precision to identify where these technologies can truly generate earnings impact, rather than piling into them because of the fear of missing out.”