Free Intensive Nov 21–22Free 2-day hands-on Vibecoding Intensive on Nov 21–22 at 16:00 MSK (Sat–Sun) · Live Cohort 3 starts Nov 23Join Free →Cohort 3 (Nov 23)
ReleaseInsight · On RadarTechCrunch→ original

Nvidia Pulls Back From OpenAI and Anthropic: What Jensen Huang Isn't Saying

Jensen Huang announced that Nvidia's investments in OpenAI and Anthropic will likely be its last. The CEO of the largest AI chip maker framed it as a push for neutrality, but analysts doubt that is the full explanation. Nvidia remains the key GPU supplier for the entire industry, and holding direct stakes in its customers created a conflict of interest. Still, regulatory risks and a strategic shift toward the company's own AI ambitions may also be behind the decision.

AI-processed from TechCrunch; edited by Hamidun News
Nvidia Pulls Back From OpenAI and Anthropic: What Jensen Huang Isn't Saying
Source: TechCrunch. Collage: Hamidun News.
Listen to article

Jensen Huang knows how to turn business decisions into theatrical gestures. On Wednesday, the head of Nvidia said that his company's investments in OpenAI and Anthropic will, in all likelihood, be its last. It sounds like a radical reversal for a company that until recently was investing actively in the main players of generative artificial intelligence. But the explanation Huang offered the public raises far more questions than it answers.

On paper, Nvidia's position looks logical. The company is the dominant supplier of graphics processors for the AI industry: its H100 and B200 chips underpin the computing infrastructure of practically every major AI lab in the world. When you sell weapons to all sides and also hold stakes in some of them, the conflict of interest stops being theoretical and becomes very real. Customers such as Google, Meta and Microsoft could reasonably ask whether OpenAI and Anthropic get preferential access to scarce GPUs precisely because Nvidia is their investor. Stepping back from direct investments eases that tension, at least on the surface.

The timeline, however, makes it hard to believe neutrality is the only reason. Nvidia took stakes in OpenAI and Anthropic when both companies were running large funding rounds and competition for access to cutting-edge chips was especially fierce. The investments looked like a strategic move then: a way to strengthen ties with key customers and get a privileged window into the development of frontier models.

What has changed? The AI chip market, though still tight, is gradually becoming more competitive. AMD is gaining share, Google is developing its own TPUs, Amazon is pushing Trainium chips, and a whole cohort of startups, from Cerebras to Groq, offers alternative architectures.

Nvidia still dominates, but its monopoly position is no longer as absolute as it was two years ago.

There is also a regulatory context that Huang chose not to emphasize. Antitrust regulators on both sides of the Atlantic are scrutinizing cross-investments in the AI sector more closely. The US Federal Trade Commission has already examined the structure of Microsoft's investment in OpenAI, and the European Commission is tightening its approach to vertical integration in the technology sector. For Nvidia, whose market capitalization exceeds two trillion dollars, even a hint of an antitrust investigation could cost tens of billions in share value. A preemptive exit from these investments is less a gesture of goodwill than insurance against regulatory risk.

There is a third hypothesis, perhaps the most intriguing. Nvidia is increasingly building its own AI capabilities, which go far beyond making chips. The company is assembling a full-cycle platform, from hardware and software to cloud services and ready-made AI solutions for enterprise customers. The DGX Cloud service, the Omniverse platform and tools for building AI agents all place Nvidia not just as a supplier but as a direct competitor of its former portfolio companies. Investing in those you plan to compete with is a bad strategy by any standard.

The market's reaction to Huang's statement was muted, which is telling in itself. Investors seem to have taken the news as an expected course correction rather than a sensation. Nvidia's shares barely reacted, which suggests analysts had already built such a scenario into their models. For OpenAI and Anthropic, losing Nvidia as an investor is not critical: both have raised enough capital from other sources, and the business relationship around chip purchases will evidently continue.

But the symbolic weight of this step is hard to overstate. Over the past three years, the generative AI ecosystem has been built on a tight web of intertwined interests: Microsoft invested in OpenAI, Google in Anthropic, Nvidia in both, Amazon in Anthropic, and so on. This web of mutual stakes created the impression of a closed club in which a few giants divide the future of the technology among themselves. Nvidia's exit from that arrangement is the first serious crack in the established structure. If others follow, the industry landscape could change fundamentally.

Jensen Huang built an empire on being indispensable to everyone at once. Giving up investments in AI labs is not a retreat but a regrouping. Nvidia is betting that selling shovels in a gold rush pays better than owning stakes in the mines. But when you start digging for gold yourself, neutrality becomes a tactic rather than a principle. And that is exactly what Huang chose to leave between the lines.

*Meta has been designated an extremist organization and is banned in the Russian Federation.

Story
ZK
Zhemal Khamidun — CPO at Alpina Digital, product manager of AlpinaGPT. Building AI products since 2019. Hamidun News is a pet project that grew into a production aggregator with 6,000+ readers.
✦ Unified AI Workspace · Nomad

Test Frontier LLMs & Autonomous Agents in One Workspace

Run Claude Opus 5.5, GPT-6, Gemini 3.1, and DeepSeek side-by-side in Nomad AI (nomadnet.ai) — unified model routing, deep research, and multimodal workflows under one subscription.

Try Nomad AI →Launch Nomad Agent
What do you think?
Loading comments…