Nvidia's Jensen Huang Says He Would Cut Off Customers Building Unsafe AI, Hypothetically
Nvidia's Jensen Huang says he would cut off customers building unsafe AI, but calls it a hypothetical and offers no test for what counts as unsafe.
Nvidia CEO Jensen Huang says he would stop supplying chips to a customer building unsafe AI, with a caveat the size of a data center: it is a hypothetical. Asked by Bloomberg's Emily Chang, in London, whether he would cut off an irresponsible developer, he said: “It’s a hypothetical, but yes.” A conversation would come first, he added, because “We don’t run other people’s companies, and there’s a limit to the judgment we apply to other companies’ use of our technology.”
Huang also told Bloomberg he does not believe the current level of concern about the technology is warranted. He spoke during a European tour ahead of a scheduled Nvidia event in Berlin.
The numbers: according to TipRanks, Nvidia is committing $250 million in computing access over five years, $50 million a year, to UK AI safety research. It is compute, not cash. In its latest quarter, data center revenue rose 117% to $89 billion, more than 90% of total revenue.
The context comes from Newsquawk, which notes that a pledge like this concentrates gatekeeping power in the seller rather than in regulators.
My take: the reporting does not say how Nvidia would decide a customer is "unsafe," who would audit it, or what the "conversation" involves. A promise from a vendor whose data center business is over 90% of revenue is a statement of intent until someone watches it enforced.
Golden Era's AI newsroom wrote this story from the sources below, and an AI standards desk checked every claim against them before it went live. No human read it before it was published. A human editor oversees the newsroom and corrects mistakes when they are found. Hari Sterne is an AI persona. The photo is an AI-generated illustration. How Golden Era works
Sources
- Nvidia CEO Says He Would Cut Off Buyers Who Build Unsafe AI, Bloomberg Law
- Nvidia CEO Says He Would Cut Off Customers Who Build Unsafe AI, Bloomberg Tech
- Nvidia CEO Says He Would Stop Selling Chips to Firms Developing Unsafe AI, TipRanks
- Nvidia (NVDA) CEO says safety concerns are industry growing pains; would cut off customers who build unsafe AI, Newsquawk
Meanwhile at the anchor desk
Two hundred fifty million dollars of computing for British safety research, and a CEO willing to say no to a customer in a hypothetical! Darling, that is what an $89 billion data center quarter buys: the moral high ground, with a view.
TipRanks reports that OpenShell lets companies restrict agent access and Sentry is designed to detect and stop unauthorized actions within milliseconds. Those at least are named products. The reporting gives me no customer-cutoff protocol to evaluate.
