OpenAI Revenue at $50bn: Why AI Stocks Fell in October

Admirals

Source published: Site published:

Original article · فارسی

OpenAI Revenue at $50bn: Why One Number Triggered an AI Selloff

Admirals
Admirals Expert Verified
Oct 12, 2026
AI Summary

OpenAI revenue was running at about $50bn a year at the end of September 2026, according to a Financial Times report confirmed by Reuters. That is roughly $18bn to $20bn below the $68bn to $70bn figure that had circulated among investors. Reports attribute the gap mainly to accounting. The higher number included gross sales made through cloud partners, which OpenAI does not book in full as its own revenue.

The news hit the AI trade on Thursday 8 October. The Nasdaq Composite fell 1.25% to 27,193.34, its worst session since mid-August, as reported by Wealth Professional. CoreWeave lost nearly 8% and Oracle about 5.5%. Nvidia slipped 2.9%. On 9 October, Bloomberg reported that OpenAI had told investors it still expects a run rate of $70bn or more by year-end, as summarised by The Next Web.

The information in this article is provided for educational purposes only and does not constitute financial advice. Consult a financial advisor before making investment decisions.

Table of Contents

This article explains how the OpenAI revenue run rate is measured and why the market reacted so sharply. It also looks at circular financing between chip suppliers and AI labs. Finally, it sets out the late-October Big Tech earnings that markets are watching next.

Key Facts at a Glance

The table below gathers the main numbers from the week. Each figure links back to its source in the sections that follow.

Item Figure Date
OpenAI run rate (FT, Reuters) About $50bn End of September 2026
Figure circulated earlier About $68bn to $70bn Late September 2026
OpenAI year-end target (Bloomberg) $70bn or more Reported 9 October 2026
OpenAI Q3 run-rate growth +77% total, +107% enterprise Q3 2026
Nasdaq Composite −1.25% to 27,193.34 8 October 2026 close
CoreWeave −7.77% to $81.58 8 October 2026 close
Oracle −5.48% to $135.69 8 October 2026 close
Nvidia −2.94% to $230.48 8 October 2026 close

Source: Reuters, The Next Web, Wealth Professional, StockAnalysis, CRWV and ORCL price history. Closes at 4:00 pm ET. Past performance is not a reliable indicator of future results. Forecasts are not a reliable indicator of future performance.

Source: TradingView - Past performance is not a reliable indicator of future results or future returns.

What Happened to AI Stocks on 8 October?

The selloff started after the Financial Times published the $50bn figure from an OpenAI investor presentation, as noted by The Next Web. The higher number had circulated in the market for more than a week. The lower figure raised a simple question. If a major AI customer earns less than thought, can it pay for all the computing power it has ordered?

Ross Mayfield, an investment strategist at Baird, told CNN that "there are going to be tremors throughout all of the related sub-industries", as quoted by Wealth Professional. Chipmakers as a group fell 3.4% that day. They had gained more than 80% so far in 2026, according to the same report.

How the Nasdaq Composite Closed

The AI selloff was concentrated in technology. The Dow Jones Industrial Average actually rose on 8 October. The next day all three main indices closed higher.

Index Close 8 Oct Change 8 Oct Close 9 Oct Change 9 Oct
Nasdaq Composite 27,193.34 −1.25% 27,366.17 +0.64%
S&P 500 7,765.36 −0.47% 7,811.54 +0.59%
Dow Jones Industrial Average 51,231.64 +0.10% 51,654.95 +0.83%

Source: Wealth Professional (8 October), AP via La Nación (9 October). Closes at 4:00 pm ET. Past performance is not a reliable indicator of future results.

Which AI Stocks Fell the Most?

Some of the heaviest losses came from companies with large OpenAI contracts. CoreWeave and Oracle sit at the leveraged end of the chain. Chip designers fell less.

Stock Close 7 Oct Close 8 Oct Change 8 Oct Close 9 Oct Change 9 Oct
CoreWeave (CRWV) $88.45 $81.58 −7.77% $82.08 +0.61%
Oracle (ORCL) $143.56 $135.69 −5.48% $141.40 +4.21%
Broadcom (AVGO) $376.51 $360.14 −4.35% $361.54 +0.39%
AMD $645.86 $620.68 −3.90% $608.10 −2.03%
Nvidia (NVDA) $237.47 $230.48 −2.94% $229.28 −0.52%

Source: StockAnalysis price history for CRWV, ORCL, AVGO, AMD and NVDA. Closes at 4:00 pm ET, changes computed from consecutive closes. Past performance is not a reliable indicator of future results.

Microsoft also traded lower during the session. By about 1:50 pm ET it was down 1.69% at $520.82, according to OfficeChai. On 9 October it rose 2.38% to $535.07 at the 4:00 pm ET close, as shown on StockAnalysis. The split between Microsoft and the chip names on Friday was notable.

Why Is OpenAI Revenue Lower Than Expected?

Based on the reports so far, the two figures mainly measure different things.

Gross and Net Revenue Accounting

OpenAI and Anthropic book partner sales in different ways, as reported by The Next Web citing Bloomberg. Anthropic counts the gross amount of some sales made through cloud partners. OpenAI counts only its own share of partner revenue.

Point OpenAI Anthropic
Sales through cloud partners Books its net share Books the gross amount of certain sales
Partners named Microsoft Amazon Web Services, Google, Microsoft
Effect on headline run rate Lower Higher
Revenue shared with Microsoft About 20% historically Not applicable

Source: The Next Web, citing Bloomberg and The Information.

Investors tried to build a like-for-like comparison with Anthropic. That calculation produced a number close to $70bn. A source familiar with the documents told CNN that the higher figure did not come from OpenAI, as noted by The Next Web. Reuters, by contrast, reported that OpenAI had indicated a figure near $70bn at a separate investor event.

How Is Run Rate Calculated?

A run rate usually takes one month of revenue and multiplies it by 12. Analysts consider it a sometimes misleading metric, as noted in the Reuters report. The figure is unaudited. It also changes with the accounting choices described above.

  • A run rate is a snapshot of one month projected over a year.
  • It can move fast when a company grows quickly or loses a large contract.
  • Private companies are under looser disclosure rules than listed ones.
  • Comparisons between AI labs may be misleading if their booking methods differ.

A Timeline of the Numbers

The table shows how the headline figure moved over two months. The August number later turned out to include Azure-related revenue that OpenAI had stopped receiving around April.

Date Report Run rate cited
Mid-August 2026 Bloomberg, The Information Above $40bn
29 September 2026 Axios Nearing $70bn
Late September 2026 CNBC source $68bn (including partner gross sales)
8 October 2026 Financial Times, Reuters About $50bn at end of September
9 October 2026 Bloomberg $70bn or more expected by end of 2026

Source: The Next Web, Reuters via The Star, Wealth Professional.

Be On The Edge

Open Live Account

Is OpenAI Revenue Growth Still Strong?

The reported growth rates remain high, although they rest on unaudited figures. OpenAI's total run rate grew 77% in the third quarter, according to a person familiar with the investor deck cited by The Next Web. Its enterprise run rate grew 107% over the same period.

Chief financial officer Sarah Friar put growth from end-June to end-September at 70%. This was reported by The Information and summarised by The Next Web. To reach $70bn by year-end from about $50bn, the run rate would need to rise by about 40% in a single quarter.

Measure Value What it implies
Run rate, end of September 2026 About $50bn Starting point
Year-end target $70bn or more Needs about +40% in Q4
Q3 growth, total +77% Above the Q4 requirement, although past growth may not repeat
Q3 growth, enterprise +107% OpenAI expects enterprise to drive most gains

Source: The Next Web, citing Bloomberg and CNBC. Past performance is not a reliable indicator of future results. Forecasts are not a reliable indicator of future performance.

OpenAI Revenue vs Anthropic

The comparison with Anthropic is what made the gap so visible. Anthropic's run rate passed $65bn at the end of July. In the second quarter its quarterly revenue overtook OpenAI's for the first time, according to Reuters as cited by The Next Web.

Metric OpenAI Anthropic
Latest run rate About $50bn (end of September 2026) Above $65bn (end of July 2026)
Q2 2026 quarterly revenue $6.7bn $11.5bn
Accounting for partner sales Net share Gross amount of certain sales
Listing plans Ruled out a listing in 2026 Could list as soon as November

Source: The Next Web, citing Reuters and Bloomberg. Past performance is not a reliable indicator of future results.

The table above shows each company on its own reporting basis. Because the methods differ, the gap may look wider than it would on a like-for-like basis. An Anthropic IPO would give Wall Street a clearer view of its finances, as noted by Reuters. This could make comparisons easier.

OpenAI Revenue Forecast and Funding Needs

OpenAI is spending heavily on computing power. It expects to burn through $278bn in cash between 2026 and 2030, according to an FT report summarised by Reuters via The Star. That is the core of the OpenAI cash burn debate.

Item Figure Source date
Projected cash burn, 2026 to 2030 $278bn 18 September 2026
Year-end 2026 run-rate target $70bn or more 9 October 2026
New funding being sought $30bn or more 9 October 2026
Pre-money valuation in talks $1.4tn 9 October 2026

Source: Reuters via The Star, The Next Web. Forecasts are not a reliable indicator of future performance.

Can OpenAI Pay for Its Compute Contracts?

This question has followed the company all year. In April 2026 The Wall Street Journal reported that OpenAI had missed internal user and revenue targets. Shares of Oracle and CoreWeave fell that day, according to Reuters via Yahoo Finance. The same report described concerns from the CFO about paying for future compute if growth slowed.

Among the OpenAI compute commitments, the Oracle contract alone is worth about $300bn, as reported by OfficeChai. A fresh funding round could ease near-term pressure. A delay or a lower valuation could revive the worries seen on 8 October.

  • OpenAI is seeking $30bn or more at a $1.4tn pre-money valuation, as reported by The Next Web.
  • OpenAI has ruled out a listing in 2026, citing safety, according to The Next Web, citing Bloomberg.
  • Several large suppliers have committed capacity that depends on OpenAI's continued spending.

Circular Financing: Who Owes Whom?

Several AI suppliers are also investors in their own customers. Critics call this circular. Money flows from a chipmaker into an AI lab, and then back to the chipmaker as orders. Elon Musk described the flows between OpenAI and its suppliers as an "infinite money glitch", as reported by OfficeChai.

Link Terms Source
Nvidia to OpenAI $30bn equity stake, not tied to deployment milestones Benzinga via Finviz
Nvidia AI equity bets in 2026 More than $40bn in total The Next Web
AMD to OpenAI 6 GW of GPU supply, warrant for up to 160m AMD shares OfficeChai
OpenAI to Oracle $300bn of computing capacity over about five years OfficeChai
OpenAI to CoreWeave Multi-year capacity contracts TIKR
Broadcom to Anthropic Loan facility of up to $42bn Pulse 2.0

Why Circularity Matters for Markets

When suppliers fund their customers, revenue across the chain depends on the same few labs. A doubt about one lab's income can therefore spread across suppliers and lenders at once. This may help explain why a single accounting report coincided with falls in shares worth trillions of dollars.

The 8 October drop appeared to reflect doubts about the numbers behind the AI supply chain. When the reported run rate came in about $20bn below the circulated figure, many companies that rely on OpenAI's future spending traded lower within hours.

There is a fair counterpoint. Equity stakes and supply deals are common in fast-growing industries. They can also help secure capacity. The risk lies in concentration, because several of these contracts rest on one loss-making customer.

Other Signals From the Same Week

Several related stories added to the mood. Each one touches the question of how AI infrastructure is financed.

  • SpaceX was reported to be seeking about $40bn of debt led by Apollo to buy Nvidia chips, according to investingLive.
  • Nvidia-backed Firmus shelved its Australian IPO after investor demand fell short, as reported by Crypto Briefing.
  • Broadcom agreed to lend Anthropic up to $42bn to help fund a long-term chip lease, according to Pulse 2.0.

Oracle and CoreWeave: The Leveraged End of the Chain

Both companies borrow heavily to build data centres for AI clients. That makes them more sensitive to doubts about customer revenue. It may also help explain why they fell the most on 8 October.

Oracle Debt Concerns

Oracle reports a large order book. Remaining performance obligations reached $664bn in its fiscal first quarter, as reported by Crypto Briefing. The OpenAI contract alone is worth about $300bn. Free cash flow was negative in the quarter because capital spending was so large.

Oracle metric Value Period
Revenue $19.3bn (+30%) Q1 FY27
Cloud infrastructure revenue $7.4bn (+121%) Q1 FY27
Remaining performance obligations $664bn Q1 FY27
Capital expenditure $28.5bn Q1 FY27
Free cash flow −$5.4bn Q1 FY27
Net debt About $132bn As listed by StockAnalysis on 12 October 2026

Source: Crypto Briefing, StockAnalysis. Past performance is not a reliable indicator of future results.

Oracle stock closed at $141.40 on 9 October after a 4.21% rebound, computed from consecutive closes before dividend adjustment. It traded in a 52-week range of $114.50 to $322.54, according to StockAnalysis. Its 2026 price change to 9 October was about −27%, as listed by CompaniesMarketCap. This figure is before taxes and costs, which may vary.

Source: TradingView - Past performance is not a reliable indicator of future results or future returns.

CoreWeave Debt Concerns

CoreWeave rents GPU capacity to AI labs and big tech firms. It funds growth with debt. Net debt stood at about $46bn, or about 8.2 times trailing EBITDA, according to TIKR.

CoreWeave metric Value
Net debt About $46bn
Net debt to EBITDA About 8.2x
Senior notes coupons 8.5% to 9.75%
Close on 9 October 2026 $82.08
8 October 2026 intraday low $80.63

Source: TIKR, StockAnalysis, StockAnalysis price history. Past performance is not a reliable indicator of future results.

CoreWeave stock has the shortest public history in this group. It listed in March 2025, so only one partial calendar year of data exists. That limits any long-term comparison.

CompaniesMarketCap lists a 2025 change of +79.03% from the March listing and +14.62% for 2026 to 9 October, before taxes and costs, which may vary. Over the 52 weeks to 9 October the price fell 41.36%, according to StockAnalysis. This is the most recent full 12-month period available for the stock. Past performance is not a reliable indicator of future results.

Source: TradingView - Past performance is not a reliable indicator of future results or future returns.

Trade with £100 on MT5

Open Live Account

Chip Designers After the Selloff

The chip designers fell less on 8 October. They have stronger balance sheets and a wider customer base. Even so, their stakes in AI labs tie part of their story to the same revenue question.

Stock Close 9 Oct Change on 8 Oct 2026 change to 9 Oct
Nvidia $229.28 −2.94% +22.94%
Broadcom $361.54 −4.35% +4.46%
AMD $608.10 −3.90% +183.95%

Source: StockAnalysis for NVDA, AVGO and AMD closes, CompaniesMarketCap for 2026 changes (Nvidia, Broadcom, AMD). Closes at 4:00 pm ET. Price changes before taxes and costs, which may vary. Past performance is not a reliable indicator of future results.

Nvidia stock kept slipping on 9 October while Oracle bounced. Analyst price targets cover a wide range. The average price target from 62 analysts stood at $328.72, with a range of $180 to $515, according to StockAnalysis. Targets are opinions and can change quickly. They are not a reliable indicator of future prices.

AMD stock fell on both days, after a strong run in 2026. That gain followed annual falls of 56.89% in 2022 and 12.84% in 2024, as the five-year table below shows. Broadcom stock slipped below levels it held before the selloff. Both companies have direct links to AI labs through supply or financing deals.

Source: TradingView - Past performance is not a reliable indicator of future results or future returns.
Analyst view Nvidia Oracle
Average price target $328.72 $237.97
Lowest target $180 $110
Highest target $515 $400

Source: StockAnalysis NVDA forecast, StockAnalysis ORCL forecast, as listed on 12 October 2026. Forecasts are not a reliable indicator of future performance.

Five-Year Performance in Context

A single bad day looks different against a longer history. The tables below cover the last five full calendar years.

Nasdaq Composite, 2021 to 2025

Year Year-end close Annual change
2021 15,644.97 +21.39%
2022 10,466.48 −33.10%
2023 15,011.35 +43.42%
2024 19,310.79 +28.64%
2025 23,241.99 +20.36%

Source: Wikipedia, Nasdaq Composite. Price index, 2021 to 2025, before taxes and costs, which may vary. Past performance is not a reliable indicator of future results.

The Nasdaq 100 followed a similar path. It rose 26.63% in 2021 and fell 32.97% in 2022, according to Slickcharts. It then gained 53.81% in 2023 and 24.88% in 2024. In 2025 it added 20.17%. These figures are before taxes and costs, which may vary. Past performance is not a reliable indicator of future results.

Wider US Market and Microsoft, 2021 to 2025

The S&P 500 and the Dow Jones Industrial Average also appear in the daily moves above. So does Microsoft, and the table shows their last five full calendar years.

Year S&P 500 (total return) Dow Jones (price) Microsoft (price)
2021 +28.71% +18.73% +55.79%
2022 −18.11% −8.78% −27.69%
2023 +26.29% +13.70% +58.35%
2024 +25.02% +12.88% +14.50%
2025 +17.88% +12.97% +16.17%

Source: Slickcharts S&P 500 returns (including reinvested dividends), Slickcharts Dow Jones returns, CompaniesMarketCap Microsoft price history. Calendar years 2021 to 2025, before taxes and costs, which may vary. Past performance is not a reliable indicator of future results.

AI Stocks, Annual Price Change 2021 to 2025

Year Nvidia Oracle AMD Broadcom
2021 +124.48% +38.91% +55.90% +61.13%
2022 −51.44% −5.39% −56.89% −13.00%
2023 +246.10% +27.84% +130.26% +106.27%
2024 +178.87% +62.10% −12.84% +116.48%
2025 +34.88% +18.36% +77.53% +50.00%

Source: CompaniesMarketCap price history for Nvidia, Oracle, AMD and Broadcom. Calendar years 2021 to 2025, price change only, before taxes and costs, which may vary. Annual changes as calculated by CompaniesMarketCap may differ slightly from other sources. Past performance is not a reliable indicator of future results.

CoreWeave listed in March 2025. Its only calendar-year figure is for the part of 2025 it traded, shown by CompaniesMarketCap, at +79.03%. The 2022 row shows how large annual declines in tech shares have been in a year of rising interest rates.

Big Tech Earnings and Capex: The Next Test

The biggest buyers of AI hardware report results in late October. Their spending plans are the demand anchor for the whole chain. Any cut in capital expenditure could matter more than the OpenAI headline.

Company Latest 2026 capex guidance Reported
Alphabet $195bn to $205bn July 2026
Amazon About $220bn July 2026
Microsoft About $175bn (calendar 2026) July 2026
Meta $130bn to $145bn July 2026

Source: W.Media (Alphabet), Stocktwits (Amazon), The Motley Fool (Microsoft, Meta). Forecasts are not a reliable indicator of future performance.

Added together, these ranges come to roughly $720bn to $745bn for 2026. Moody's expects hyperscaler capex to approach $1tn in 2027, as reported by DCD. Amazon chief executive Andy Jassy said the company would still lack the capacity to meet all 2026 demand, according to Stocktwits.

Earnings Calendar to Watch

Microsoft and Alphabet are expected to report on 28 October, according to StockAnalysis listings for Microsoft and Alphabet. AMD follows on 3 November, as listed on StockAnalysis. Meta and Amazon usually report in the same week as Microsoft. Their dates had yet to be confirmed on the pages checked for this article.

Date Event Why it matters
27 to 28 October FOMC meeting Rate path and long-term yields
28 October Microsoft and Alphabet results Capex plans and cloud growth
Late October (to be confirmed) Meta and Amazon results Capex plans and AWS growth
3 November AMD results Progress on OpenAI GPU deliveries
Late November (to be confirmed) Nvidia results Data centre demand

Source: Federal Reserve calendar, StockAnalysis. Dates may change, and the economic calendar lists the latest schedule.

Alphabet's report gets a closer look in our Alphabet Q3 2026 earnings preview. For Meta, see the Meta stock analysis for October 2026.

Macro Backdrop: Yields and the Fed

The AI story did not unfold in calm conditions. Long-term US yields were near 24-year highs and oil was above $100 a barrel. Markets were also pricing a further Federal Reserve hike.

Indicator Level Note
US 10-year Treasury yield About 5.23% Near a 24-year high
US 30-year Treasury yield About 5.60% Near a 24-year high
Brent crude $104.28 +4.07% on 8 October
WTI crude $91.49 +3.64% on 8 October
Chance of a December Fed hike 69.2% CME FedWatch via Reuters

Source: Wealth Professional, data from 8 October 2026, end of session (WTI at settlement). Past performance is not a reliable indicator of future results.

The Fed, led by Chair Kevin Warsh, raised rates in September. Minutes from that meeting showed most officials saw another hike as likely by year-end, according to Scotsman Guide. Higher yields tend to weigh on growth stocks, because they reduce the present value of distant profits.

The next US inflation reports are listed on the BLS release schedule. A hotter reading could add pressure on tech valuations. A softer one could ease it.

AI Stocks Outlook: Scenarios Traders Watch

The scenarios below are illustrative only. They describe possible market reactions to the triggers discussed in this article. They are not predictions.

Scenario Possible triggers How markets could react
Positive Hyperscalers confirm or raise capex, OpenAI closes its funding round Tech indices could retest recent highs
Neutral Capex holds steady with little upside surprise Trading could stay choppy, and cash-rich firms could hold up better than leveraged ones
Negative A capex cut, a delayed OpenAI round or a weak Anthropic IPO Leveraged names such as Oracle and CoreWeave could come under renewed pressure
Negative (macro) Hot inflation data or a Fed hike on 28 October Higher yields could weigh on the whole growth sector

Illustrative scenarios based on the triggers discussed in this article. Forecasts are not a reliable indicator of future performance.

Is AI a Bubble? Arguments on Both Sides

This question splits analysts. The table sets out both views. Neither side can be proven right in advance.

Arguments for strong demand Arguments for caution
OpenAI's enterprise run rate grew 107% in Q3 Run-rate figures are unaudited and can be misleading
Amazon says it cannot meet all 2026 demand OpenAI projects a cash burn of $278bn to 2030
Big Tech capex is still rising Suppliers finance their own customers
Anthropic and OpenAI revenue keeps growing fast Oracle and CoreWeave carry heavy debt

Source: The Next Web, Stocktwits, Reuters via The Star, TIKR.

The mood on 8 October showed how quickly sentiment can turn. It also showed how fast part of the move can reverse. For background on the private company itself, see our guide to OpenAI stock and its private status.

Invest with £1

Open Live Account

How Traders Often Approach an AI Selloff

During a stock market selloff, traders often focus on a few practical questions. These are observations of common market behaviour. They are not recommendations.

  • Traders often separate cash-rich companies from those that rely on debt.
  • Many watch the economic calendar closely in the weeks of major earnings.
  • Some reduce position sizes when volatility rises.
  • Others wait for the next set of hard data before forming a view.
  • Some traders give more weight to Big Tech spending plans than to reports about private companies.

Ways to Gain Exposure

Admirals offers CFDs on shares of US companies such as Nvidia and Oracle through its share CFD range. It also offers CFDs on indices, including a US100 index CFD. CFDs let traders speculate on rising or falling prices without owning the underlying asset.

CFDs are complex instruments and come with a high risk of losing money rapidly due to leverage. You should consider whether you understand how CFDs work and whether you can afford to take the high risk of losing your money.

New traders can explore the platforms with a demo account. A demo account uses virtual funds, and results may differ from live trading. For a primer on the product, read what CFD trading is.

Risk Management Checklist

Leverage magnifies both gains and losses. Volatile sectors such as AI can move several per cent in a day.

Step Purpose
Setting a stop-loss level before opening a position Helps limit losses if the market moves against a position
Sizing positions relative to account equity Helps manage risk across several trades
Checking the calendar for earnings and data Helps identify scheduled events that may cause price gaps
Reviewing margin requirements Helps in monitoring the risk of margin close-outs

General educational checklist. A stop-loss may not protect against gaps or slippage.

Our guide to risk management covers these tools in more detail. Our Nvidia stock forecast page also tracks analyst views on the largest AI chipmaker.

Conclusion

Based on the reports so far, the OpenAI headline appears to be mainly an accounting story. A figure built on gross partner sales met a figure built on net sales, and the gap was about $20bn. Markets appeared to react because so much of the AI chain depends on OpenAI's ability to keep growing and keep paying.

The next tests are the Fed meeting and Big Tech earnings in late October. Spending plans from Microsoft and Alphabet may tell investors more than any private run rate. Meta and Amazon are expected to report soon after, with dates yet to be confirmed.

Frequently Asked Questions

What is OpenAI's revenue run rate in 2026?

OpenAI's annualised revenue run rate was about $50bn at the end of September 2026, according to Reuters.

Why was the $68bn figure higher?

Reports say it included gross sales through cloud partners. OpenAI books only its net share of those sales.

Why did AI stocks fall on 8 October 2026?

Reports suggest investors worried that lower OpenAI revenue could weaken its ability to pay for large computing contracts.

Which stocks fell the most in the AI selloff?

CoreWeave fell about 7.8% and Oracle about 5.5%. The large chip designers recorded smaller declines that day.

What is circular financing in AI?

It describes suppliers investing in AI labs that then buy the suppliers' chips or cloud capacity with that money.

When do Microsoft and Alphabet report earnings?

Both companies are expected to report on 28 October 2026, according to StockAnalysis listings checked on 12 October.

Is the AI bubble about to burst?

Nobody can know in advance. Strong demand data sits alongside heavy debt, so outcomes could differ widely.

Some of the content on this page was created with the help of automated tools. All final text has been reviewed by Admirals and is under its editorial responsibility.

INFORMATION ABOUT ANALYTICAL MATERIALS:

The given data provides additional information regarding all analysis, estimates, prognosis, forecasts, market reviews, weekly outlooks or other similar assessments or information (hereinafter “Analysis”) published on the websites of Admirals investment firms operating under the Admirals trademark (hereinafter “Admirals”) Before making any investment decisions please pay close attention to the following:

  • This is a marketing communication. The content is published for informative purposes only and is in no way to be construed as investment advice or recommendation. It has not been prepared in accordance with legal requirements designed to promote the independence of investment research, and that it is not subject to any prohibition on dealing ahead of the dissemination of investment research.
  • Any investment decision is made by each client alone whereas Admirals shall not be responsible for any loss or damage arising from any such decision, whether or not based on the content.
  • With view to protecting the interests of our clients and the objectivity of the Analysis, Admirals has established relevant internal procedures for prevention and management of conflicts of interest.
  • The Analysis is prepared by an analyst (hereinafter “Author”). The Author Admirals is an employee /a contractor for Admirals. This content is a marketing communication and does not constitute independent financial research.
  • Whilst every reasonable effort is taken to ensure that all sources of the content are reliable and that all information is presented, as much as possible, in an understandable, timely, precise and complete manner, Admirals does not guarantee the accuracy or completeness of any information contained within the Analysis.
  • Any kind of past or modelled performance of financial instruments indicated within the content should not be construed as an express or implied promise, guarantee or implication by Admirals for any future performance. The value of the financial instrument may both increase and decrease and the preservation of the asset value is not guaranteed.
  • Leveraged products (including contracts for difference) are speculative in nature and may result in losses or profit. Before you start trading, please ensure that you fully understand the risks involved.

Sources

Share this post
Admirals An all-in-one solution for spending, investing, and managing your money

More than a broker, Admirals is a financial hub, offering a wide range of financial products and services. We make it possible to approach personal finance through an all-in-one solution for investing, spending, and managing money.