The widespread adoption of Artificial Intelligence is expected to cause significant disruption across many industry sectors, and markets are already pricing in the potential impacts. But is it too soon to identify the winners and losers in the AI revolution?
Read this article to understand:
- Why markets are now looking for losers as well as winners from the roll-out of AI
- While some stocks now look cheap, they may struggle to recover very quickly
- Why it makes sense to screen securities for the risk of AI disruption
Global equities have rebounded sharply in recent weeks as hopes of an end to hostilities in the Middle East and robust corporate earnings from several leading US companies propelled shares to record highs.
The MSCI World index, having on March 30 been down 6.3 per cent for the year, has since surged 14.6 per cent. The rise in global markets has been paced by the US as the rally in technology shares shows little sign of running out of steam amid an ongoing boom in spending on Artificial Intelligence (AI). The S&P 500 index hit a record high on May 11.1
However, behind this seemingly rosy picture a more turbulent series of events has been playing out. While the boom in AI spending may be floating an ever-expanding universe of stocks, a new chapter in the story is being written. The central theme: a bifurcation in the market as investors scramble to identify not just the winners but also the potential losers from the new technology.
The Software & IT Services sectors have been at the epicentre of the fallout. Dubbed by some the “SaaSpocalypse”, shares have been hit by fears generative AI will disrupt traditional software-as-a-service (SaaS) business models.
Whereas the S&P 500 index has returned 8.1 per cent the year to date, the IT services and data processing sectors have each lost more than a quarter of their value, with software not far behind.
The sell-off has been fuelled by fears AI agents will replace per-seat software licenses, make software development cheaper and increase competition, and threaten incumbent SaaS companies with obsolescence. There are a number of channels via which firms could be hurt as AI is rolled out.
The four horsemen of the SaaSpocalypse
Aviva Investors equity analyst Josep Bori says the rapid development of systems that can autonomously execute complex workflows, has raised fears much of the software industry is going to be destroyed.
“AI is raising question marks about the demand for traditional software. However, it’s important to recognise not all software companies will be impacted equally,” he says.
Bori believes those companies selling software applications that do not provide a system of record or data repository are among the most vulnerable. Records and data represent a stored value or essential data for users, without which a software application might seem more dispensable.
Nonetheless, while the deployment of AI could threaten some companies’ business models, fears AI might make a whole raft of software products redundant look wide of the mark, at least for now.
Despite rapid progress, AI remains fundamentally a non-deterministic technology. Unlike say a traditional calculator, which follows a rigid, predictable path to the same result every time, it relies on elements of randomness and probability which means identical instructions or queries may result in different results.
Moreover, it is prone to ‘hallucinate’, in other words to generate a response that contains false or misleading information and yet is presented as fact. While the industry is still debating if hallucinations can be eliminated, Bori believes this unreliability is unacceptable for many business use cases and implies traditional software code will continue to dominate.
This is especially true of areas such as cyber security, infrastructure software and engineering software. The level of sophistication involved makes it hard to see how they could be threatened by AI.
Indeed, in the case of cyber security, it is possible to envisage AI providing a boon to software providers, since bad actors could make use of the technology to devise increasingly sophisticated cyber attacks and scams.
Wider market impacts
The AI fallout has not been confined to software and IT services. Recent weeks have seen severe declines in share prices across multiple other sectors including media, advertising, publishing and insurance. All were prompted by fears the rollout of AI could create havoc for companies.
Figure 1: AI disruption, year-to-date return
Source: Bloomberg, as at 13 May 2026.
The slide in shares has followed a series of announcements from US based AI group Anthropic about its AI assistant Claude. They have suggested it has progressed rapidly from being a chatbot to a system that can do real work in legal, sales, finance, data analysis, coding and security.
That has left investors grappling with how to value companies across multiple sectors whose competitive positioning could look very different in a few years.
Edward Kevis, equity fund manager at Aviva Investors, says the magnitude and speculative nature of the market moves is complicating the task of separating winners from losers when making investment decisions.
“Panic is setting in whenever a headline pops up about Anthropic. At this stage it’s all highly speculative. We’ve seen some modest downgrades in analysts’ earnings forecasts but nowhere near what would be needed to justify the extent of the share price falls,” he says.
The slide in asset prices has not been confined to equities: bond markets have been impacted too.
Aviva Investors’ head of investment-grade credit, James Vokins, says the extra yield offered by some corporate bonds relative to comparable government debt is as much as ten per cent higher than might otherwise be expected. Like Kevis, he says market movements are being driven more by speculation than actual evidence companies are being affected.
Nonetheless, he says it makes sense to take a cautious approach by screening companies and even sectors for the risk they will struggle to survive the impact of AI.
He has avoided bonds issued by companies with business models that have not been tested over the long term. Many lie within technology sectors, but the list also includes companies in travel and leisure. He is also looking for extra yield in other sectors he fears will be disrupted.
"But only at the margin, as at this stage it is inconclusive what the impact will be. Many of the companies we're investing in are going to benefit from AI as it will make them more efficient,” Vokins says.
Ultimately, the scale of investment in AI will only be justified if computers begin replacing humans, and primarily service sector workers, in huge numbers. Such fears have resulted in real estate shares being hammered too.
However, Kevis argues the market is pricing in such huge disruption to certain industries such as software and IT services, that it is hard to see how authorities will not then feel obliged to intervene to limit AI’s impact.
“By implication, the market is taking a very aggressive view on how much earnings and growth is going to slow for some of these businesses. Ultimately, most countries operate service-sector based economies. It’s open to debate how much job displacement will be tolerated,” he says.
As for Bori, he believes indiscriminate selling has left many software stocks looking cheap. For instance, shares in Workday, the US-based Human Resources software provider, have plunged more than 40 per cent since the start of the year on fears the company’s business will be disrupted by Anthropic.
Yet paradoxically, as Bori points out, Anthropic uses Workday to help manage its own Human Resources function.
“This goes to show that if it makes sense for a company to develop a software system internally they might leverage AI to do it. But not everything is going to be built from scratch,” he says.
That said, while the sell-off may appear overdone, both he and Kevis say it is unclear betting against the market in the hope of making a quick return would be wise.
“Even though we think software has been probably too broadly and too much impacted, the concern we have is that there is no clear catalyst for the market to change its mind,” Bori says.
All the same, Kevis is in no rush to sell out of positions that have been impacted, without clearer evidence AI is eroding revenues and profitability.
Microsoft shares, which have fallen 13.8 per cent year to date on fears AI could damage sales of the company’s Office software, are a case in point.
The fall looks hard to rationalise when considering its big data-centre and cloud-computing businesses mean the company is participating in the ‘AI trade’ and should allow it to continue growing earnings strongly even if its software business is affected.
Widely held across the equity funds he manages, Kevis says he is extremely comfortable owning the shares at current levels.
“When you look at metrics such as free cashflow yield and price/earnings ratios, these valuations are attractive for the financial profile of the shares,” he says.
Few doubt that AI will have both positive and negative effects across industry sectors. There will be both winners and losers. But with AI technology moving a pace, the scale and precise form of those effects are still extremely uncertain.
“In a world where AI has the potential to disrupt almost every issuer and business model, the discipline lies in remaining agile while applying clear selection guardrails, so uncertainty broadens our thinking without diluting conviction," says Vokins.
References
- Source for all figures is Bloomberg, as at 13 May 2026.
