CoreWeave posts $3.66 billion in revenue as AI debt risks grow

AI investment is drawing both market enthusiasm and warnings about financing risk. CoreWeave reported third-quarter revenue of $3.66 billion, up 168%, and a contracted backlog of $132.5 billion, while carrying nearly $47 billion in debt. Some of its bonds recently yielded more than 16%. Finance chief Nitin Agrawal said the company does not plan to issue more unsecured parent-level debt through 2027 and instead plans to sell up to 35 million shares. Its results lifted shares 16% in U.S. premarket trading, with Nasdaq futures up 1.6% and S&P 500 futures up 0.7% on Wednesday. Nvidia shares also rebounded after a steep decline.

The rally came amid debate over interest rates and whether AI spending will earn adequate returns. October core CPI rose 0.3%, above the 0.2% forecast. Federal Reserve Governor Michelle Bowman warned that stocks could fall sharply if AI investments disappoint and estimated AI-related debt could exceed $800 billion this year. Christopher Waller raised the possibility of a half-point rate increase if inflation stays hot, while Dallas Fed President Lorie Logan called for at least two more quarter-point hikes. Higher borrowing costs could also weigh on housing and small-business investment. Moody’s Atsi Sheth warned that refinancing older debt may become more expensive next year.

Nvidia’s financing ties are part of regulators’ broader scrutiny of AI-related credit risks. The company has made major commitments to AI infrastructure firms, including cloud-service purchases and project guarantees. Its shares fell 30% from an October 6 peak through Tuesday’s close before gaining more than 4% by midday Wednesday. Experts said even much larger Nvidia losses would probably not cause the kind of system-wide crisis regulators seek to prevent, but falling GPU values could leave borrowers unable to repay debt. Separately, Seagate says long-term agreements reserve most of its supply for data centers through 2028. It expects December-quarter revenue of $4.8 billion, give or take $100 million, but has not set 2028 prices. Its fiscal Q1 2027 operating margin was 48.9%, up from 26.9% a year earlier; 81% of revenue came from data centers, leaving results exposed to shifts in demand and pricing.

Anthropic’s planned Nasdaq listing, reported at a $1.4 trillion valuation, was on investors’ radar. The CoreWeave coverage also identified Anthropic as a customer planning to debut next week. In India, concern about an AI bubble has pressured companies connected to data centers, power and cooling: 34 of 42 firms Goldman Sachs identified as AI-related fell an average of 7% in October, while Anant Raj and Himadri Speciality Chemical each dropped 25%. India’s investment case includes a projected $250 billion data-center boom, but returns remain unproven, and the country lacks a homegrown AI product competing at the scale of ChatGPT, Claude or Gemini.

AI’s practical uses and limitations featured in several other reports. A survey of multifamily executives found 75% said AI improved operations, with 60% reporting greater efficiency and 55% better decisions. Yet 40% encountered new challenges, including data-quality problems and continued staff-training needs. Researchers comparing evaluation methods for two analytics agents found trace-based testing delivered the strongest balance of precision and recall on 72 known risk cases, with a 0.90 F1 score. Text-only testing matched recall but generated five times as many false alarms. The study found 24 new failures, none shared between functional testing and security red-teaming, and recommends using both approaches.

In healthcare, the Electronic Frontier Foundation sued for records about WISeR, a federal program launched in January that uses AI to review Medicare requests for certain procedures. EFF says records point to improper denials, operational problems and lengthy delays that leave patients waiting in pain. Meanwhile, connected-device security firm Finite State raised $27 million for international expansion and autonomous features. Its work helps manufacturers identify discrepancies between planned software and what runs in finished devices, amid requirements including the EU Cyber Resilience Act and FDA premarket cybersecurity rules. The IoT security market is projected to reach $80.3 billion by 2031.

Two companies also announced funding for specialized AI products. StitchStudio raised $8 million to expand tools for employee benefits, life and annuity, and commercial property and casualty insurance. Its platform automates work across older systems while leaving recommendations and client decisions to employees; the company says proposal turnaround fell from 90 minutes to 15 and customers have saved more than 5,000 hours. Separately, OKX Global General Manager Lennix discussed the potential of AI trading agents and tokenization, emphasizing risk management as these technologies develop.

Key Takeaways

  • CoreWeave posted third-quarter revenue of $3.66 billion, up 168%, and a $132.5 billion contracted backlog, but has nearly $47 billion in debt; some bonds yielded more than 16%.
  • CoreWeave plans to avoid new unsecured parent-level debt through 2027 and sell up to 35 million shares; its results helped lift its shares 16% in premarket trading.
  • Federal Reserve officials warned of inflation and AI investment risks: October core CPI rose 0.3%, Bowman estimated AI-related debt could exceed $800 billion this year, and Waller discussed a possible 50-basis-point hike.
  • Nvidia’s commitments to AI infrastructure companies are under scrutiny; experts flag potential borrower losses if GPU values fall, while saying a Nvidia decline alone is unlikely to cause a system-wide crisis.
  • Seagate has reserved most supply for data-center customers through 2028 and expects December-quarter revenue of $4.8 billion, but has not set 2028 pricing; data centers account for 81% of its revenue.
  • Anthropic’s planned Nasdaq listing was reported at a $1.4 trillion valuation; Indian AI-linked stocks fell amid doubts about returns, and India lacks a homegrown product competing at the scale of ChatGPT, Claude or Gemini.
  • A multifamily survey found 75% of executives said AI improved operations, while 40% reported challenges such as data quality and staff training.
  • Research found trace-based AI-agent evaluation achieved a 0.90 F1 score on 72 known risk cases; functional testing and security red-teaming exposed separate failures, supporting use of both.
  • EFF sued for records on Medicare’s WISeR AI review program, citing concerns over denials, operational problems and delays.
  • Finite State raised $27 million for connected-device security; StitchStudio raised $8 million for insurance AI and reports cutting proposal turnaround from 90 minutes to 15. OKX’s Lennix discussed AI trading agents, tokenization and risk management.

Higher Rates Put CoreWeave’s AI Debt Under Pressure

The 10-year U.S. Treasury yield rose above 5% as Federal Reserve officials discussed faster rate hikes, while CoreWeave reported nearly $47 billion in debt. CoreWeave’s third-quarter revenue rose 168% to $3.66 billion, and its contracted backlog reached $132.5 billion, but some of its bonds recently yielded more than 16%. Finance chief Nitin Agrawal said the company does not plan to issue more unsecured parent-level debt through 2027, and it plans to sell up to 35 million shares instead. Moody’s Atsi Sheth said borrowers refinancing older debt may face higher costs next year, while Fed Governor Christopher Waller raised the possibility of a half-point hike. The article also notes that about $650 billion in commercial property debt comes due in 2027 and that CoreWeave customer Anthropic plans to debut on Nasdaq next week.

Stocks Rise as CoreWeave Results Renew AI Optimism

Stocks gained on Wednesday as investors reacted to strong results from AI infrastructure company CoreWeave, even as expectations for Federal Reserve rate hikes increased. Nasdaq futures rose 1.6% and S&P 500 futures gained 0.7%, while CoreWeave shares climbed 16% in U.S. premarket trading. European and Asian markets also advanced, with technology shares leading gains in Asia. Investors watched Anthropic’s planned listing at a reported $1.4 trillion valuation, while Nvidia shares rose after recent steep losses. U.S. inflation data showed core prices rose 0.3% in October, above the 0.2% forecast, and several Fed officials supported further rate increases. The report also covered moves in bond yields, currencies, gold, and oil.

Fed Governor Warns AI Investment Could Put Stock Values at Risk

Federal Reserve Governor Michelle Bowman warned that U.S. stock prices could fall sharply if AI investments do not deliver expected returns. She also said higher long-term borrowing costs could hurt housing and small-business investment, and that persistent inflation may support another rate hike in December. October core CPI rose 0.3% from the previous month, above expectations. Governor Christopher Waller said the Fed should consider a 50-basis-point increase if inflation stays hot, while Dallas Fed President Lorie Logan said at least two more quarter-point hikes were needed. Bowman estimated that AI-related debt could exceed $800 billion this year and warned that off-balance-sheet financing and financial links between large tech firms and smaller companies could spread risk when markets come under pressure.

Survey Finds AI Boosts Multifamily Operations but Brings Challenges

A survey of multifamily executives found that 75% of respondents said AI improved their operations. Sixty percent reported greater efficiency, and 55% said AI helped them make better decisions. The survey also found that 70% saw improvement in data-driven decisions, with 60% citing better forecasting and 55% better risk management. However, 40% reported new challenges, including data quality problems and the need for ongoing staff training. Respondents said training and stronger data checks can help teams use AI effectively.

EFF Seeks Records on Medicare’s AI Review Program

The federal government launched WISeR in January to use AI when reviewing Medicare beneficiaries’ requests for certain procedures. The Electronic Frontier Foundation sued for records about the program, seeking more transparency. In an EFFector podcast episode, EFF’s Lena Cohen discusses records that the group says document improper denials, operational problems, and lengthy delays. The delays have left patients waiting in pain, according to the article. The episode is available on major podcast platforms.

Study finds trace based tests best at evaluating enterprise AI agents

Researchers compared three evaluation frameworks on two analytics agents using frozen execution traces. In repeated tests, 97% of configurations met the stability threshold, regardless of the agent. On 72 known risk cases, trace based evaluation had the best balance of precision and recall, with an F1 score of 0.90. Text only evaluation matched its recall but produced five times more false alarms. The study found 24 new failures, with no overlap between functional testing and security red teaming, and recommends using both approaches.

Finite State raises $27 million to expand connected device security

Finite State raised $27 million to expand its global operations, strengthen its presence in international markets, and develop autonomous features for its security platform. The company helps manufacturers identify differences between planned software and what actually runs in finished devices, and find risks across software supply chains. New rules, including the EU Cyber Resilience Act, FDA premarket cybersecurity requirements, UNECE R155 and R156, and IEC 62443, require manufacturers to show that connected products are secure. Finite State serves customers in areas such as power grids, cars, medical devices, homes, hospitals, and factories. The IoT security market is projected to reach $80.3 billion by 2031.

US AI stock losses raise questions about Indian companies

A month of losses in US technology stocks has raised concerns about an AI bubble and affected Indian companies linked to data centres, power, and cooling. Of 42 Indian companies Goldman Sachs identified as AI related in September, 34 fell by an average of 7% in October. Anant Raj and Himadri Speciality Chemical each dropped 25%, while Bharti Airtel and Blue Star were among the few major gainers. Investors are questioning whether AI can earn enough to justify the huge investments, and the Nasdaq fell 5% in October. India’s AI investment case depends in part on a projected $250 billion data centre boom, but the article says returns remain unproven and India lacks a homegrown AI product that competes at the scale of ChatGPT, Claude, or Gemini.

OKX leader discusses AI trading agents and risk management

OKX Global General Manager Lennix discussed crypto trading in an interview on The Sujal Show. He highlighted the potential of AI trading agents and the importance of managing risk as the technology develops. The conversation also covered tokenization and its possible role in changing financial markets by improving access and liquidity. The article says traders should watch how AI and tokenization affect trading, while noting that this content is not financial advice.

Nvidia's AI financing links raise concerns about wider financial risks

Nvidia has made major financial commitments to AI infrastructure companies, including cloud service purchases and project guarantees. The Federal Reserve and other financial institutions are examining whether debt and connected financing could spread losses if the AI market weakens. Nvidia shares fell 30% from their October 6 peak through Tuesday's close, then rose more than 4% by midday Wednesday. Experts who track AI financing say even much larger Nvidia losses would probably not cause the kind of system wide crisis regulators seek to prevent. However, a fall in GPU values could leave companies that borrowed to buy them unable to repay debts, while technology companies face both credit problems and falling stock prices.

Seagate’s AI hardware demand is booked through 2028, but pricing remains uncertain

Seagate says long-term agreements through 2028 reserve most of its supply for data center customers. It expects December quarter revenue of $4.8 billion, plus or minus $100 million, after reporting $4.2 billion in the September quarter, but has not set prices for 2028. Seagate’s operating margin reached 48.9% in fiscal Q1 2027, up from 26.9% a year earlier, as better pricing and product mix outweighed some higher costs. Compared with Apple, Seagate has a lower earnings multiple and faster recent revenue growth, but 81% of its revenue came from data centers, and its recent margin gains could weaken if prices or demand change. Its next report is expected in January 2027.

StitchStudio raises $8 million to expand its insurance AI products

StitchStudio raised $8 million to expand its AI products for employee benefits, life and annuity, and commercial property and casualty insurance. Investors included Lookout Ventures and Grit Capital Partners, alongside continued investment from ManchesterStory and BrokerTech Fund and support from insurance veterans. The company helps insurers and brokers automate work across older, disconnected systems and varying rules while keeping employees responsible for recommendations and client decisions. Customers including M3 Insurance, Heffernan Insurance Brokers, Fidelity Life, InstaBrain, and Frank Winston Crum Insurance use the platform for tasks such as quote comparison, producer onboarding, eligibility, billing reconciliation, and benefits quoting. StitchStudio says proposal turnaround dropped from 90 minutes to 15 minutes, with more than 5,000 hours saved.

Sources

NOTE:

This news brief was generated using AI technology (including, but not limited to, Google Gemini API, Llama, Grok, and Mistral) from aggregated news articles, with minimal to no human editing/review. It is provided for informational purposes only and may contain inaccuracies or biases. This is not financial, investment, or professional advice. If you have any questions or concerns, please verify all information with the linked original articles in the Sources section below.

AI investment AI infrastructure CoreWeave AI debt Nvidia interest rates AI bubble Anthropic data centers AI agent evaluation AI security testing healthcare AI Medicare IoT cybersecurity insurance AI AI trading agents tokenization Seagate

Comments

Loading...