• NVIDIA (NVDA)'s planned $100 billion investment in OpenAI has been dramatically scaled back to approximately $20 billion amid renegotiations.
  • CEO Jensen Huang publicly denies "drama" while private concerns about OpenAI's strategy and competitors surface.
  • The partnership is evolving into a more traditional vendor relationship as OpenAI diversifies its hardware suppliers.

A Partnership in Flux

NVIDIA CEO Jensen Huang has repeatedly dismissed reports of friction with OpenAI, most recently declaring that "there's no drama involved" and the partnership remains "on track." However, behind the public reassurances, the companies are renegotiating the terms of their strategic relationship, with NVIDIA's actual investment likely shrinking from the initially announced $100 billion to around $20 billion, according to people familiar with the matter.

In September 2025, NVIDIA and OpenAI announced what appeared to be a landmark partnership: NVIDIA would invest up to $100 billion in OpenAI and help build 10 gigawatts of AI infrastructure. The first gigawatt was targeted to come online in the second half of 2026 using NVIDIA's Vera Rubin platform. But by late January 2026, The Wall Street Journal reported friction between the companies, citing Huang's private criticism of OpenAI's business strategy and concerns about competitors like Anthropic (ANTH) and Google (GOOGL). The report indicated the deal was nonbinding and that discussions had shifted toward a significantly reduced equity investment.

Market Reactions and Industry Shifts

Huang called these reports "nonsense," insisting NVIDIA would "definitely participate" in OpenAI's funding round. Yet current talks suggest NVIDIA's contribution to OpenAI's early-2026 funding round is approximately $20 billion—substantially lower than the original headline figure. This reduced stake occurs alongside major investments from Amazon (AMZN) (potentially $50 billion) and SoftBank (SFTBY) (up to $30 billion) in a total funding round targeting $100 billion. NVIDIA's stock tumbled 3.4% following investor anxiety about the deal's status, reflecting broader market unease.

The partnership restructuring mirrors industry-wide shifts as competitors like Google and Anthropic increasingly adopt custom ASICs instead of relying solely on NVIDIA GPUs. OpenAI is reportedly exploring alternative inference chip providers due to performance concerns, signaling a move away from single-vendor dependence. "What we're seeing is a natural evolution in a maturing market," said one industry analyst who requested anonymity. "NVIDIA is no longer the only game in town."

Future Implications

Efforts to maintain the partnership have hit a snag, with the collaboration now appearing to evolve into a more traditional vendor relationship rather than the transformative joint infrastructure project initially envisioned. Without a robust deal, NVIDIA risks losing its dominant position in OpenAI's supply chain. The first phase of infrastructure deployment remains targeted for the second half of 2026, though under modified partnership terms that emphasize flexibility over exclusivity.

Attempts to reach NVIDIA and OpenAI for additional comments were unsuccessful, but sources indicate both companies remain publicly committed to collaboration despite the scaled-back financial involvement. The situation highlights how even high-profile tech partnerships must adapt to competitive pressures and strategic realignments in the fast-moving AI sector.