
What happened
Nvidia is reducing its financial guarantee for OpenAI's data center project from the originally planned $250 billion to just under $120 billion, covering the first construction phase (about five gigawatts of capacity). Investor concerns about Nvidia's risk exposure drove the cutback, while OpenAI separately negotiates a lease for the full 10-gigawatt project and Nvidia talks separate financing for up to $350 billion in chip purchases.
Why it matters
The scaled-back guarantee signals caution even among the biggest beneficiaries of the AI boom, giving ammunition to critics warning of an AI bubble. Conversely, Anthropic's financials push the opposite narrative: the company's revenue more than doubled from $4.73 billion in Q1 to over $11.5 billion in Q2 (a 14x year-over-year increase), and projects $190 billion to $200 billion in annual revenue for 2028—suggesting sustained, strong demand for proprietary AI services amid competition and political headwinds.
What to watch
Anthropic reportedly plans to go public at a valuation near $1 trillion in late September or early October. A financial services firm (Ramp) recently measured a slight flattening in demand for Anthropic tokens among business customers, a metric worth monitoring against the company's growth projections.
Summaries like this, in your inbox every morning.
Nvidia's decision to halve its guarantee reflects mounting investor scrutiny of mega-scale AI infrastructure spending, even as the company and its partners pursue unprecedented data center buildouts. The original $250 billion commitment represented an extraordinary concentration of risk on a single project; the reduction to just under $120 billion suggests that Wall Street is demanding more conservative exposure despite the obvious strategic importance of securing capacity for OpenAI. The guarantee covers only the initial phase, with OpenAI and Nvidia pursuing separate financing for the full scope—a structure that may indicate risk-sharing across multiple parties rather than a single actor bearing the full downside.
Anthropric's financial trajectory, by contrast, suggests the investment community's caution is not uniform across all AI bets. A 14x year-over-year revenue jump and projections of $190 billion–$200 billion in 2028 revenue indicate that proprietary AI services are capturing real, durable demand. That growth rate—more than 10x annually in each of the three years leading up to early 2026—outpaces typical venture-scale software growth and points to a market expanding faster than skeptics anticipated. The slight flattening in token demand among Ramp's business customers is a note of caution, but it has not yet translated into a revenue deceleration in the company's public filings.
Pick your industry and the AI tools you use, and get news related to your work every day.
Free · 30 seconds with Google · unsubscribe anytimeWhat is AIToday? →
Ask AI anything about this article. The AI reads this article, earlier AIToday articles, and Wikipedia, and cites its sources. Q&As are published on this page for other readers too.
Supermicro Computer appointed Mark Long as chief financial officer, bringing an experienced technology finance…

Holtek Semiconductor unveiled its 2026 product roadmap on October 15, focusing on four core areas: smart prote…

TSMC reported about NT$763.1b in Q3 net profit and confirmed that demand for advanced processors is running ah…

Nvidia-backed CoreWeave steadied Thursday after a sharp slump, as investors worried about the debt financing b…

Weathernews said on October 15 it implemented its own AI weather-forecasting model, AI-Nowcast, in its Weather…

TOPPAN Holdings and Keio University ran a quantum-computer and generative-AI proof-of-concept from February 1…
