The artificial intelligence boom has been defined by extraordinary technological breakthroughs, record-breaking corporate valuations and an unprecedented race to build the computing infrastructure needed to power the next generation of digital innovation. Yet recent market turbulence has exposed a more fragile reality. An analysis published by The Observer UK argues that while AI remains one of the world’s most transformative technologies, a convergence of geopolitical tensions, financial pressures and growing leverage has prompted investors to question whether the industry’s rapid expansion is creating new systemic risks.
The latest concerns emerged in late June when US Commerce Secretary Howard Lutnick sought a meeting with executives from ASML, the Dutch manufacturer whose highly specialised lithography equipment sits at the centre of global semiconductor production. Lutnick reportedly wanted clarification over concerns that one of the company’s extreme-ultraviolet (EUV) lithography machines had reached China in breach of US-led export restrictions.
ASML firmly rejected the allegation, stating that it knows the precise location of each of its 340 EUV lithography systems and that none are operating in China. Nevertheless, the concerns reflected broader anxieties surrounding technological competition between Washington and Beijing. Those worries intensified when reports surfaced that an unnamed Shanghai-based company had developed its own, albeit less advanced, lithography machine and planned to manufacture five units next year.
The reports reverberated across financial markets. ASML’s shares briefly declined, while South Korea’s technology-heavy Kospi index extended a dramatic sell-off that had already erased more than 40% of its value over the previous five weeks before staging an 18% rebound on Friday as institutional investors stepped in to buy discounted shares. The sharp swings underscored the sensitivity of global technology markets to developments within the semiconductor supply chain.
The repercussions extended far beyond listed companies. Goldman Sachs estimates that as many as 360,000 retail trading accounts have been wiped out during the recent market turmoil, while more than one million South Korean investors—roughly one in every 30 working-age people—have received margin calls requiring them either to deposit additional funds or liquidate investments. Highly leveraged hedge funds in Western markets have also suffered significant losses.
According to The Observer UK, China’s reported progress in lithography technology is only one factor unsettling investors. Confidence has also been dented by the emergence of the open-source Chinese artificial intelligence model Kimi K3 and the high-profile stock market listing of Chinese chipmaker CXMT. Collectively, these developments have challenged assumptions about continued US dominance in advanced artificial intelligence and semiconductor technology.
Alongside geopolitical competition, investors are confronting another concern: the enormous sums being committed by major technology companies to finance AI infrastructure. As data centre construction accelerates, questions are increasingly being asked about both the scale of corporate borrowing and the sustainability of the financing models supporting the industry’s expansion.
Those concerns have been amplified by changes in the broader financial environment. US government borrowing costs climbed to their highest levels in almost two decades after the Federal Reserve decided to leave interest rates unchanged. Thirty-year Treasury yields rose as investors demanded higher returns amid concerns over inflation, volatile energy prices and declining US strategic oil reserves. Markets were also unsettled by the communication approach of the Federal Reserve’s new chair, Kevin Warsh, whose limited public explanations of policy decisions added to uncertainty in Treasury markets.
Higher borrowing costs are expected to tighten financing conditions across the corporate sector, particularly for hyperscale technology companies that rely heavily on debt to fund AI investments. Sebastian Raedler of Bank of America observed that increasing financing costs can turn projects that would otherwise be economically viable into unviable investments.
The scale of borrowing illustrates the industry’s dependence on financial markets. Amazon, Alphabet, Meta and Oracle collectively issued approximately $194bn in bonds during 2026 up to 7 July, representing a 79% increase compared with the same period a year earlier. These companies have already committed more than $1tn to artificial intelligence capital expenditure.
Corporate performance, however, has not been uniform. Amazon and Microsoft continued to record robust cloud revenue growth, whereas Meta’s shares declined after investors questioned the company’s rapidly increasing AI spending without a clearly defined path towards future revenues. Meanwhile, investors have demanded higher premiums to hold debt issued by hyperscalers, with credit spreads widening significantly.
Andrew Keches, a managing director at Barclays, described the scale of recent bond market movements as highly unusual, saying they were “a few-times-in-career-type moves” that were reminiscent of the Covid period. At the same time, the cost of protecting against corporate defaults has risen sharply. Prices for credit default swaps linked to major technology companies have reached record levels in recent weeks. Oracle’s five-year credit default swap, for example, rose from 144 basis points at the beginning of the year to 215 basis points by Wednesday, increasing the annual cost of insuring $10m of debt to approximately $215,000.
Alen Lin, a senior director at Fitch Ratings, attributed part of Oracle’s financial pressure to the scale of infrastructure it is building for contracted customers such as OpenAI and Meta, arguing that the company lacks sufficient internal cash generation to finance those commitments independently.
The Observer UK also highlights concerns about increasingly interconnected financing arrangements across the AI ecosystem. In analysis cited from the Bank for International Settlements, Phurichai Rungcharoenkitkul argues that the combination of rapid expansion, extensive borrowing and circular investment structures raises important questions about long-term sustainability and financial stability.
Nvidia provides a notable example of those interconnected relationships. The chipmaker guarantees approximately $3.5bn of loans taken out by certain customers and has reportedly considered extending similar financial support to OpenAI for a $250bn data centre project, according to the Wall Street Journal. Such arrangements illustrate how financial exposures increasingly extend across companies throughout the AI supply chain.
The analysis warns that these connections could amplify future shocks. Because hyperscalers including Google, Oracle and Microsoft hold equity stakes in AI laboratories such as Anthropic and OpenAI, financial difficulties at a single AI developer could spread rapidly through corporate balance sheets, potentially forcing investment cutbacks elsewhere and transmitting stress throughout the wider technology ecosystem. As Andrew Keches observed, the principal “tail risk scenario” centres on AI laboratories themselves, with any indication that they are unable to meet financial obligations representing “a material shift in the narrative”.
Although the United Kingdom occupies a comparatively modest position within the global AI supply chain and the FTSE 100 is regarded by some investors as an “anti-tech” index, the analysis concludes that Britain would not escape the effects of a major global technology correction. The Bank of England has warned that any substantial reassessment of AI company valuations could significantly affect the UK economy. In a scenario where US equity markets declined by 45% following weaker expectations for AI productivity and profitability, the Bank estimates the UK economy would contract by 2.2%.
For now, investment in artificial intelligence continues at remarkable pace. Yet, as The Observer UK notes, scepticism is steadily increasing. According to S3 Partners, investors betting against Amazon, Microsoft and Meta have increased their short positions by more than $20bn this year. Recent events have demonstrated that while the AI revolution continues to advance, turbulence affecting its largest participants can reverberate across financial markets, national economies and the broader global investment landscape.
