What happened?
On 12 November 2026 Nvidia, the company whose chips power most of the world’s AI, announced that Jensen Huang had resigned as president, chief executive and director, effective immediately. He co-founded the company in 1993. The chief financial officer, Colette Kress, is interim chief executive while the board looks for a permanent successor, and Harvey Jones is the board’s first chair. Nvidia gave no reason. Its shares fell nearly 10% that day.
Does this change the AI tools we use?
Not today. Your organisation almost certainly does not buy anything from Nvidia directly. You use chat assistants, translation tools and cloud services sold by other companies, and the announcement, as the commentary reports it, says nothing about their prices or terms.
The reason to pay attention is the argument made in a commentary by Tianze Zhang on the AIFOD website. Zhang writes about governments in developing countries, not about NGOs. But his subject is dependence on a single supplier, and that is a problem small organisations know well.
What is the argument?
Zhang’s point is that the Global South came to depend on one company for AI hardware without choosing to, and that the company depended on one person more than most people realised. He thinks Nvidia is still strong, but expects a difficult period while it has only an interim chief. In such periods, he writes, small buyers wait longest and pay most.
He also sees an opening. If Nvidia is less dominant, more suppliers compete for the same customers, and customers get better terms on price, local capacity and training.
What can an NGO take from it?
Zhang gives governments three steps and one change of focus. Here is how we would translate them for a non-profit. The translation is ours, not his.
- Ask in writing. His first step is to review every contract and partnership in progress and ask the supplier in writing whether dates and prices still hold. If your work depends on a technology grant, a non-profit discount or donated cloud credits, find the end date and the renewal terms now, and ask the provider to confirm them by email.
- Keep a second option. He tells governments to write tenders that more than one platform can win, and asks universities to teach skills that carry across platforms. For an NGO that means not building a whole workflow around one tool. Keep your prompts, templates and data in formats you can move, and train staff in the task, not only in the product.
- Buy together. He advises countries to purchase through regional bodies, because a group gets terms that a small buyer cannot. NGO networks and umbrella bodies can do the same when they ask vendors for discounts or training.
- Judge AI by what it does. Zhang wants attention to move from how many chips a country owns to what is done with them: local data, local talent, and uses in health, farming, education and public services that run on efficient, open models. For an NGO the question is the same. Which task got better, and for whom?
What this page is, and is not
This page summarises one member’s commentary. NGO Service is run by AIFOD, which also published that commentary and says member commentaries do not necessarily reflect its position. Zhang does not discuss why Huang left, and we have not either. Nothing here is a prediction about any product you use. The full text is “After Jensen Huang: What Nvidia’s Leadership Change Means for Developing Countries”.