Investors betting against US cloud computing and data centre companies have accumulated approximately $6 billion in unrealised profits during October 2026, as growing concerns about debt-financed artificial intelligence infrastructure investments trigger a sharp decline in share prices. The selloff has intensified scrutiny of companies racing to expand computing capacity to meet rising AI demand, with investors questioning whether the cost of building and maintaining these facilities can be sustained.
The companies facing pressure include CoreWeave, IREN, Nebius Group, Applied Digital and Core Scientific. These businesses operate in the rapidly expanding market for computing infrastructure used by artificial intelligence developers and other technology companies. Their expansion plans require substantial investment in data centres, servers and related infrastructure, increasing concerns about financing costs and the risks associated with heavy borrowing.
Data analytics firm ORTEX reported that short interest in the five companies ranged from approximately 20% to 26% of their publicly available shares, also known as free float. Such elevated levels indicate that a significant proportion of tradable shares have been borrowed and sold by investors expecting prices to decline.
When the share price falls, investors can potentially profit from the difference, although the strategy also carries the risk of losses if prices rise instead.
CoreWeave has become a major focus of bearish trading activity. Short interest in the company reached 23.7% of its free float, up from 19.6% at the beginning of October. The figure was also above the 22.7% recorded at the start of 2026.
The company’s shares have fallen almost 30% during October, generating nearly $1.7 billion in unrealised profits for investors betting against the stock. These gains remain on paper unless the positions are closed and the profits are realised.
The pressure on CoreWeave reflects broader concerns about the financial demands of expanding AI computing infrastructure. Companies providing specialised cloud services must invest heavily to secure equipment and build capacity, while their financial performance depends on generating sufficient revenue from customers to justify those expenditures.
Nebius Group has produced even larger paper gains for short sellers during the month. Bearish investors have accumulated approximately $2.94 billion in unrealised profits from positions in the company so far in October.
However, the monthly gains contrast with the broader performance of those trades throughout 2026. Short sellers betting against Nebius have recorded an unrealised loss of $3.33 billion year to date, demonstrating how movements in the company’s share price have created substantial opportunities and risks for investors taking opposing positions.
Short interest in Nebius has risen to 20% of its free float, compared with 15.4% at the beginning of the year. The increase suggests that bearish investors have expanded their positions as concerns surrounding AI infrastructure spending and financing have grown.
Other data centre operators have also experienced significant short-selling activity. Investors betting against IREN and Applied Digital have generated combined unrealised profits exceeding $1.2 billion during October, according to the reported figures.
Core Scientific, another company operating in the data centre sector, had short interest equivalent to 23.1% of its free float. Investors taking bearish positions against the company recorded approximately $115 million in paper profits during the month.
Together, the five companies have around $17 billion worth of shares sold short, ORTEX reported. The firm also indicated that borrowing shares remained relatively inexpensive for all five businesses, making it easier for investors to maintain their short positions.
The latest figures underline the extent of bearish positioning across a group of companies closely associated with the expansion of artificial intelligence computing. Although demand for AI services has encouraged substantial investment in infrastructure, the scale of spending has also raised questions about how quickly companies can translate that demand into sustainable financial returns.
Building AI data centres requires significant capital, including spending on advanced computing equipment, electricity infrastructure and facilities capable of supporting energy-intensive workloads. Companies that finance expansion through borrowing must also manage repayment obligations and other financing costs while attempting to increase revenue.
If investors become less confident that future earnings will justify these commitments, share prices can come under pressure. Falling valuations may, in turn, intensify concerns about the financial risks facing businesses that depend on continued investment and access to capital.
Trading on Thursday reflected the negative sentiment surrounding the sector. CoreWeave and Nebius shares fell more than 4% in early trading. IREN declined 6.1%, while Applied Digital lost 3.7% and Core Scientific dropped 2.4%.
The declines added to the gains accumulated by short sellers during October, although the figures do not necessarily represent profits that have already been secured. Unrealised returns can change quickly if share prices recover, and short sellers may face losses if the market moves against their expectations.
The latest market developments highlight the growing divide between enthusiasm for artificial intelligence and concerns about the financial structures supporting its expansion. Demand for computing capacity has encouraged companies to pursue ambitious development plans, but investors are increasingly examining the debt, capital requirements and potential returns associated with those investments.
For short sellers, the recent decline has created substantial paper gains. For the companies involved, however, sustained pressure on their share prices could make investor confidence and future financing more important as they continue developing infrastructure for the AI industry.
The market’s direction will depend partly on whether these businesses can demonstrate that their investments generate sufficient revenue and earnings to support their expansion strategies. Until investors gain greater confidence in the financial sustainability of the sector, highly leveraged AI infrastructure companies may remain vulnerable to further volatility.
