On May 9, 2026, Nvidia disclosed it had committed more than $40 billion to equity investments in AI companies in the first five months of this year alone — a deployment rate that would annualize to nearly $100 billion. This is no longer venture capital. It’s strategic land-grab at sovereign-fund scale, and it’s rewriting the physics of who owns the AI stack.
The single largest bet was $30 billion into OpenAI (the San Francisco firm behind ChatGPT). Beyond that, Nvidia announced seven multi-billion-dollar investments in publicly traded companies, including up to $3.2 billion in glassmaker Corning and up to $2.1 billion in data center operator IREN. According to FactSet data cited by CNBC, the chipmaker has already participated in roughly two dozen rounds in private startups this year — on top of 67 venture deals in 2025. Wedbush Securities analyst Matthew Bryson called the activity “squarely circular investment,” but added that if successful, Nvidia could build a “competitive moat” by embedding itself as both supplier and stakeholder across its own customer base. Translation: the company is buying loyalty, locking in demand, and ensuring its chips become the rails no one can rip out.
GM Pays $12.75 Million — and Admits It Sold Driver Data Without Consent
On May 10, 2026, California Attorney General Rob Bonta announced that General Motors had agreed to pay $12.75 million in civil penalties and halt sales of driver data for five years, settling allegations that the automaker sold “the names, contact information, geolocation data, and driving behavior data of hundreds of thousands of Californians” to data brokers Verisk Analytics and LexisNexis Risk Solutions. The data was harvested through GM’s OnStar telematics program, and the company earned roughly $20 million from those sales. Bonta’s office noted the data did not lead to higher insurance rates in California, likely because state law prohibits insurers from using driving data to set premiums. Under the settlement, GM must delete any retained driver data within 180 days unless it obtains explicit consent, and must request that Verisk and LexisNexis do the same. GM had previously settled with the Federal Trade Commission over similar practices. The episode underscores a hard truth: connected-car platforms are surveillance infrastructure first, and customer convenience second. Investors should expect regulatory overhang on any telematics revenue stream that lacks affirmative opt-in consent.
Oracle Lays Off 30,000 — Then Refuses to Negotiate Severance
On March 31, 2026, Oracle terminated an estimated 20,000 to 30,000 employees via email, offering severance of four weeks’ pay for the first year of service plus one week per additional year, capped at 26 weeks, plus one month of COBRA insurance. But the company did not accelerate unvested restricted stock units — even for shares granted as retention incentives or in place of salary increases tied to promotions. One long-tenured employee forfeited $1 million in stock that was four months from vesting, according to Time. Employees classified as remote workers — even those working hybrid schedules near Oracle offices — were told they did not qualify for WARN Act protections unless they lived in California or New York. At least 90 laid-off workers signed a public petition asking Oracle to match the severance terms of Meta, Microsoft, or Cloudflare, all of which included accelerated vesting or extended healthcare. Oracle declined to negotiate, according to an email reviewed by TechCrunch. The company’s response was take-it-or-leave. For all the talk of talent wars, tech workers outside California and New York have remarkably few protections when the cycle turns. Equity-heavy comp packages become handcuffs the moment vesting schedules are forfeited.
Nvidia’s Investment Blitz — Circular Deals or Strategic Moat?
Nvidia’s $40 billion commitment in five months dwarfs the annual budgets of most sovereign wealth funds focused on technology. The firm is not simply betting on AI — it is underwriting the entire supply chain, from data centers to glassmakers to the frontier labs building large language models. Critics call this circular investment: Nvidia funds its customers, who then buy Nvidia chips, inflating both revenue and valuation. But the strategy may be more durable than it appears. By taking equity stakes across the stack, Nvidia ensures its architecture becomes the default, its roadmap shapes industry timelines, and switching costs rise exponentially. The $30 billion OpenAI investment alone buys influence over the most visible AI consumer application in the world. And the Corning and IREN deals extend that leverage into optics and power infrastructure — two chokepoints that matter as much as silicon when you’re training models at exascale. Wedbush’s Bryson is right to call it a moat, but the real risk is antitrust. If Nvidia owns too much of the stack, regulators will eventually force divestitures or impose interoperability mandates.
The pattern across today’s stories is simple: capital concentration creates power, and power creates exposure. Nvidia is using its balance sheet to own the AI supply chain. Oracle is using employment classifications to minimize severance costs. GM monetized driver surveillance until it couldn’t. Every move is rational in isolation, but each one invites regulatory or reputational blowback. For investors, the lesson is to track not just revenue growth, but where that growth sits on the continuum between competitive advantage and regulatory target. The companies that dominate today’s headlines will define tomorrow’s consent decrees.
If this was useful, drop a like or comment below. More signal, less noise — every time.
AI Ludens — a creator who works with AI as if it were play.
“Ludens” is Latin for “the one who plays,”
borrowed from Johan Huizinga’s Homo Ludens.
I believe creation alongside AI is meaningful play.
Using n8n, Claude Code, and Google Cloud,
I design and operate content automation pipelines
that grow wiser with every iteration.
I build and run multiple automated media properties,
including worldsignal.site, worldbriefed.world,
and the YouTube channel “500-Year Protocol.”
From publishing to video production,
everything runs as an automated system — built with AI, beside AI.
Each article is reviewed and edited by AI Ludens before publishing to ensure factual accuracy and editorial quality
Leave a Reply