Nvidia pulls back from OpenAI, Anthropic investments
Nvidia is done writing checks to OpenAI and Anthropic
At the Morgan Stanley Technology, Media and Telecom conference in San Francisco on March 4, Nvidia CEO Jensen Huang said the company’s $30 billion investment in OpenAI “might be the last time” it invests in the AI startup. He made the same statement about Nvidia’s $10 billion stake in Anthropic.
The reason: both companies are expected to go public later this year, and once the IPO window opens, private investment deals of this kind typically close. Huang put it simply — a $100 billion follow-up investment in OpenAI is “probably not in the cards.”
For small business owners who rely on AI tools built by these companies, the shift from private funding to public markets is worth understanding. It will affect how these companies make money — and eventually, what you pay.
What’s actually happening behind the scenes
The surface explanation is clean: IPOs close the door on private mega-deals. But TechCrunch reported that the reality is more complicated.
The circular money problem. Nvidia invests billions in OpenAI. OpenAI spends billions buying Nvidia chips. MIT’s Michael Cusumano called the arrangement “kind of a wash.” OpenAI’s February 2026 funding round raised $110 billion total — $30 billion from Nvidia, $50 billion from Amazon, and $30 billion from SoftBank. Much of that capital flows right back to Nvidia through GPU purchases. Market observers have drawn bubble comparisons.
Nvidia is becoming a competitor. The company is moving beyond selling hardware into AI cloud and enterprise services. When Nvidia launches products that compete with OpenAI’s API or Anthropic’s enterprise contracts, being a major investor in both creates an uncomfortable conflict.
Diplomatic tension with Anthropic. Two months after Nvidia announced its $10 billion investment, Anthropic CEO Dario Amodei publicly compared U.S. chip companies selling processors to approved Chinese customers to “selling nuclear weapons to North Korea” — without naming Nvidia directly. The relationship has gotten complicated fast.
What Nvidia is really doing is stepping out of the business of picking sides. It sells GPUs to OpenAI, Anthropic, xAI, Google, and every other company racing toward the frontier. The arms dealer doesn’t get to have a favorite army.
Why the AI IPO wave matters for small businesses
Both OpenAI and Anthropic are expected to go public in late 2026 or early 2027. Together with SpaceX, these listings could put roughly $2.9 trillion in new market cap on public exchanges.
Here’s why that matters to a business with ten employees:
The subsidy era has an expiration date
Right now, AI tools are priced below cost. OpenAI is projected to burn $14 billion in 2026. Anthropic’s margins have improved dramatically — swinging from negative 94% in 2024 to roughly positive 40% in 2025 — but the pressure to grow revenue is constant.
Private investors tolerate losses in exchange for growth. Public investors demand profitability. When OpenAI and Anthropic answer to shareholders instead of venture capitalists, the pressure to raise prices or introduce advertising will increase. Axios drew a direct parallel to the “millennial lifestyle subsidy” that gave us cheap Uber rides and DoorDash deliveries — services that got meaningfully more expensive once those companies went public.
Competition keeps prices from spiking overnight
The good news: neither company operates in a vacuum. Google DeepMind, Meta’s Llama models, and a growing roster of open-source alternatives ensure that if one provider raises prices aggressively, customers have somewhere else to go.
Aggregate token pricing — the cost of generating text with AI — has continued to fall. At equivalent performance levels, AI inference costs drop roughly 10x per year. A query that cost $20 per million tokens in late 2022 costs about $0.40 today. Competition and better hardware are working in your favor.
The two-speed market is forming
OpenAI and Anthropic are already splitting into different business models. OpenAI is pursuing advertising to subsidize consumer AI. Anthropic is positioning itself as the premium, safety-focused enterprise option — Anthropic’s annualized revenue hit $20 billion at the Morgan Stanley conference, with paid enterprise customers driving most of that growth.
For small businesses, this means you’ll likely see cheaper, ad-supported AI tools on one end and higher-quality, privacy-focused tools at a premium on the other. The middle ground may thin out.
What to do as an AI tool buyer
You don’t need to panic about AI pricing. But you should be strategic about how you adopt and pay for these tools.
-
Avoid over-reliance on a single provider. If your entire workflow depends on one AI company, a price change could hit hard. Diversify where practical. Our guide on how to evaluate AI tools walks through the criteria that matter.
-
Lock in annual pricing where it makes sense. If you’re on a tool that offers annual plans at current rates, that’s a hedge against post-IPO price increases. Just make sure the tool is delivering real value first.
-
Watch for open-source alternatives. As we covered in our Anthropic funding analysis, the flood of capital into AI research is producing more efficient, smaller models. Open-source options are becoming viable for specific tasks — especially scheduling, intake, and basic customer service.
-
Track what you’re actually spending on AI. If you haven’t recently audited your AI budget, now is a good time. Know what you’re paying, what value it delivers, and where you’d cut if costs rose 20%.
What to watch in 2026
- IPO filings from OpenAI and Anthropic. The S-1 documents will reveal their actual margins and growth trajectory. That’s when you’ll know how sustainable current pricing really is.
- NVIDIA GTC 2026 kicks off March 16 in San Jose. Jensen Huang’s keynote will likely address the next generation of AI infrastructure — Rubin GPUs and updates to their agentic AI platform — which directly affects future tool costs.
- Subscription price changes. If a tool you use announces a price increase, evaluate whether the value still justifies the cost. Don’t pay more out of habit.
The bottom line
Nvidia stepping back from OpenAI and Anthropic isn’t a sign that AI is slowing down. It’s a sign that the industry is maturing — moving from venture-backed experimentation to public-market accountability. For small businesses, that transition will bring more pressure on pricing but also more competition, more choices, and better tools.
The practical move is to stay diversified, stay informed, and build your AI strategy around value rather than loyalty to any single provider. If you’re not sure where your business stands, talk to our team — we help businesses across Appalachia navigate exactly these kinds of shifts.