Bloomberg Professional Services
Artificial intelligence (AI) dominates everything from corporate strategy to conference panels and every major firm claims to be investing in it. But, when you look beyond the claims, how much progress is really being made?
“All the financial companies that I and my colleagues follow say they’re using AI and investing in AI, but spending overall will remain flat, so I wanted to know where they plan to cut,” said Kevin Ryan, senior insurance analyst at Bloomberg Intelligence, introducing new research at September’s Bloomberg Investment Management Summit.
Companies generally are under pressure to show technological ambition, but many asset managers are cautious when it comes to investing in and implementing AI, with many skeptical of its near-term potential.
PRODUCT MENTIONS
To better understand their thinking, Bloomberg Intelligence (BI) commissioned the Asset Managers & AI: BI Survey, an independent study of 100 CIOs and CTOs from the world’s largest asset managers. Conducted in the spring of 2025, the survey asked how close rhetoric comes to reality when it comes to AI spend.
The headline finding was that although AI features prominently in conversations, it is not yet necessarily reflected in spending because expenditure isn’t supported by evidence of significantly higher revenues. But that doesn’t necessarily imply complacency, more likely a cautious approach and a healthy dose of realism.
Caution on tech spend
According to survey results, around 70% of firms spend less than 10% of their operating expenses on technology, while for three-in-ten respondents that figure is between 10% and 20%. In EMEA, 67% of IT budgets go towards maintaining existing tech, leaving limited headroom for innovation projects. New investment flows mainly into cloud computing and data analytics, areas considered proven and regulator-friendly.
“Most of the budget goes on maintaining legacy systems,” said Ryan. “You can’t buy a new bit of software if it doesn’t segue with what you’ve already got – that would be a waste of money – and if you’re pushing two bits of software together, then there’s a process involved.”
Budget outlooks suggest limited change in the near term. The survey indicates that just over half (54%) of respondents expect technology spending to rise by less than 5% in the next year. Only 4% plan increases north of 10% – and none of these were based in Europe. Asked about their motivations for tech spending, respondents’ top answer was upgrading the business, followed by optimizing costs.
However, asked about the cost impact of AI, most respondents express modest expectations. Roughly three-quarters say it will integrate gradually into existing processes without a major effect on spending. A surprisingly low number, just 5%, believe it will materially reduce expenses within three to five years.
Ryan said: “Given the challenging operating environment, the focus is going to be on cutting costs because revenue gains are hard to come by…it’s really all about trying to get value for money.”
Ryan noted that figures likely reflect practical circumstances on the ground, not an ideological resistance to change. Questions to answer first involve where the money for any AI spending will come from and the safety of such investments in highly regulated environments.
Why the gap? Risk, regulation, cultural inertia
Asset managers are custodians of client capital and operate under strict fiduciary duty. Deploying probabilistic or opaque models into investment decisions raises difficult questions about explainability and accountability. Until standards for model governance and transparency improve, it’s natural to want to wait rather than lead.
“There is a significant compliance issue with deploying AI in some parts of your business, and that needs careful consideration,” said Ryan. “The asset management industry is well placed to benefit by not getting sucked into spending a lot of money on something that might not do too much good.”
Even so, there are signs of change. Asset managers report using AI to summarize research, detect sentiment in analyst commentary and test investment hypotheses (Bloomberg). These initiatives are primarily experiments, but they hint at a shift to come.
In coming years, as AI tools become more transparent and governance in this area develops into something more comprehensive and robust, the industry’s comfort level will likely adjust in kind.
What happens if asset managers ignore AI?
For firms on the sidelines, the consequences may not be immediate, but long-term lagging is possible.
Over the next decade, AI should change what efficiency, analysis and client service look like to asset managers. One view among market participants is that early experimentation may offer optionality if tech and regulatory landscapes mature, with the trade-off of costs and actual execution. Wait and see is a viable strategy, but it might come with longer-term risks.
Regardless of whether AI turns out to be the industry’s panacea, there could also be implications for talent acquisition and retention down the line. Sought-after technologists and data scientists increasingly want to work where AI is a priority.
And, on the client side, some investors are preferring faster, more interactive information delivery (Bloomberg) – something AI does well.
But weighed against all this, for respondents to the survey, is the upfront expense of AI deployment and doubts about its ability to boost profits. “AI isn’t that helpful in a cost-cutting environment and doesn’t necessarily, right now, help optimize performance,” said Ryan.
In a previous Bloomberg global banking survey, published in November 2024, 60% of respondents said they would cut headcount within three to five years, with 24% saying they expected between a 5% and 10% reduction. But in the Asset Managers & AI: BI Survey, only 10% of asset managers answered the question at all (many citing privacy concerns). Of those that did, half believed AI impact would increase, not reduce, employment.
In the meantime, however, some competitors are making moves. Banks, fintechs and tech-forward asset managers are embedding AI across risk management, client engagement and even trading.
That doesn’t mean firms should rush headlong into untested tools: “Will the asset management industry get left behind? Not if everybody’s doing the same thing,” said Ryan. “There’s no outlier or big spender saying that this is the one thing that’s going to change our jobs and make everything rosy in the garden.”
The case for AI confidence
Across markets, AI is moving from theory to adoption and from adoption to application. Firms are gradually finding where it adds value. Progress is incremental, but progress is happening all the same, and as confidence grows, the gap between talk and action is beginning to narrow.