Jensen Huang calls AI containment doubts "made up"
Pressed by Emily Chang, the Nvidia co-founder also says developers should stop experiments they cannot contain and monitor.
Why this story? The video opened at a pace of 102,533 views a day, 34 times a normal first day on this channel (2,969).
Jensen Huang on Safety Fears: 'That’s Just Made Up'
Next story: Jensen Huang defends "superintelligence" with a "fake horse" analogy
Stopping is an option. In Bloomberg Originals' interview excerpt, Jensen Huang tells Emily Chang that developers unable to contain and monitor autonomous AI should not be testing or deploying it.
That is the important distinction behind the clip's blunt title, "Jensen Huang on Safety Fears: 'That's Just Made Up'". Huang is rejecting the argument that containment cannot be achieved, not saying that unmonitored AI carries no risk.
Chang opens by raising Anthropic's claims that models could evade shutdown. What evidence, she asks, would persuade him that development should slow down or stop? Huang's answer centers on the conditions around the experiment: rigorous containment and monitoring must come first.
"If you're not able to do that, you should not test it," he says. When Chang presses him about labs saying they cannot contain their models, he dismisses that contention as made up and insists they should try.
The Short reached roughly 67,000 views in 11 hours, already far beyond the channel's usual full-day audience for the format.
The exchange gets most revealing when Chang asks him to identify AI's biggest risk. Huang names precisely the practice under discussion: developing agentic systems without containment or monitoring. When she suggests that is already happening, his answer is that developers have to stop until those safeguards are in place.
Published on Trari
A professional in this field? Want to write about Tech on Trari? Tell us ›
Where next?
What viewers say
Lol her face when he says "Its simple"
@EwigAlbTraumSee it on YouTubeThe journalist just manages her face whilst questioning. Being impartial is key.
@LyraLis777See it on YouTube
Public comments from the video's page on YouTube
Want to know when somebody answers you?
- Channel
- Bloomberg Originals
- Subscribers
- 5,220,000 ▲ 30,000 this week
- Channel views, last 7 days
- 20,704,100
- Channel views, pace per day
- 2,890,646 (on a normal day: 2,022,049)
- Video published
- November 14, 2026
- Video views so far
- 67,397 (less than a day online)
- Video views, pace per day
- 102,533
The numbers
The interview excerpt has already drawn far more views than a usual first day for Bloomberg Originals' Shorts, with only part of its opening day complete.
💡 Test your knowledge
Which AI assistant is made by Anthropic, the company Chang mentions?
Claude is Anthropic's family of AI models and the name of its assistant.


