Artificial intelligence (AI) has shifted from market darling to market threat in a matter of weeks, unsettling investors and knocking billions off companies once considered insulated from technological change.
For much of the past three years, AI was a euphemism for growth.
Fund managers crowded into technology giants and software providers on the promise that automation would lift profits.
Now the same logic is being turned on its head: if AI can perform high-value tasks once done by people, someone’s revenue is at risk.
The selling began in software.
The iShares Expanded Tech-Software Sector ETF, a fund that tracks United States software stocks, has fallen sharply this year, with only a handful of its constituents in positive territory.
Valuations, measured by the forward price-to-earnings ratio - a gauge comparing share prices to expected profits - have compressed to levels last seen more than a decade ago.
From there, the anxiety spread.
Insurance brokers were hit after a Madrid start-up, Tuio, launched a ChatGPT-based insurance app.
Investors questioned whether AI tools could bypass traditional intermediaries, the brokers who sit between customers and insurers to arrange cover.
Shares in Marsh McLennan and Arthur J. Gallagher fell heavily in a single session.
Analysts were quick to argue the reaction was excessive.
Large commercial policies, they noted, require complex underwriting and personal advice.
Indeed, while AI may assist with that work, replacing it is another matter.
Wealth managers were next.
When fintech firm Altruist unveiled an AI tool capable of automating tax planning - structuring investments to legally reduce tax - investors marked down traditional brokers, including Charles Schwab and LPL Financial.
The fear is unambiguous: if software can generate financial plans at low cost, fee margins could narrow.
Commercial property groups then found themselves in the firing line.
Shares in CBRE Group, Jones Lang LaSalle and Cushman & Wakefield posted double-digit falls across two days.
The concern is not only that AI might automate brokerage tasks, but that widespread job displacement could shrink demand for office space over time.
Logistics stocks delivered one of the sharpest reactions.
A small company, Algorhythm Holdings - formerly known for karaoke machines - promoted an AI freight optimisation tool designed to match loads and routes more efficiently.
Optimisation in this context means using algorithms, or sets of coded instructions, to reduce empty truck miles and fuel costs.
Shares in freight operator RXO and logistics group C.H. Robinson Worldwide tumbled, even as bond markets signalled little concern about their long-term solvency.
Market strategists describe the current mood as “shoot first, ask questions later”.
Simply put, investors are selling on headlines before earnings are affected.
Credit spreads - the extra yield investors demand to hold corporate bonds - have remained relatively steady, suggesting lenders see less immediate danger than equity traders.
Private credit firms have also been scrutinised.
Many have lent heavily to software companies.
If AI undermines those borrowers, default rates could rise.
UBS has modelled a severe scenario in which U.S. private credit defaults climb into double digits, though that remains hypothetical.
Beyond the U.S., political clarity has begun to matter more as capital rotates.
In Japan, the election victory of Prime Minister Sanae Takaichi’s party has been read as policy continuity, lifting the Nikkei 225 to fresh highs.
Thailand’s ruling party secured a decisive mandate, prompting a rally in the SET Index.
Indonesia moved the other way after MSCI flagged concerns about market transparency, triggering a sharp fall in the Jakarta Composite Index.
The broader question is whether the AI investment boom can justify the scale of spending underway.
According to UBS, capital expenditure by large U.S. technology groups - money spent on data centres and infrastructure - is set to absorb almost all operating cash flow this year.
For now, markets appear caught between two impulses: fear that AI will erode profits across white-collar industries, and belief that the technology will ultimately enhance productivity.
As always, share prices are moving faster than fundamentals, and whether that proves prophetic or premature will depend not on headlines, but on earnings.
