Nvidia CEO says 'we have to shut the labs down' if AI experiments are unsafe
- ID
- 28181
- Status
- summarized
- Published
- 24 Sep 2026, 11:24 PM
- Fetched
- 25 Sep 2026, 12:40 AM
- Provider
- Tom's Hardware
- Category
- technology
- Original URL
- https://www.tomshardware.com/tech-industry/big-tech/nvidia-ceo-says-we-have-to-shut-the-labs-down-if-ai-experiments-are-unsafe-jensen-huang-says-frontier-ai-lab-fears-are-a-distraction-not-a-call-for-regulation
- Source URL
- https://www.tomshardware.com/feeds/all
Summary
- Score
- 4.0
- Created
- 25 Sep 2026, 12:40 AM
- Tags
- Audience
- developersai_ml_learnersai_agent_usersfounders
What happened
Nvidia CEO Jensen Huang said that if AI experiments are unsafe, 'we have to shut the labs down,' pointing to civil and criminal liability over rogue agents. In the same remarks he called frontier AI lab fears a 'distraction' rather than a call for regulation. The article text available provides no specifics on what 'unsafe' means, which labs, what enforcement mechanism, or any timeline or proposed policy.
Why it matters
The concrete claim here is that liability for rogue agents would land on whoever runs the experiment — not on a new regulator — and the article gives no definition of 'unsafe' or who would enforce it. If you ship agents that take real actions (payments, file writes, outbound messages), that framing means the burden of proof sits with you, so the decision to make now is what your agent is allowed to do without a human in the loop, not whether some future rule will cover you. There is no Malaysia- or SEA-specific angle in the supplied text, so treat this as background positioning rather than a local policy signal.
Discussion angle
Huang frames liability as the enforcement mechanism instead of regulation — ask the room: if an agent you deployed caused a financial loss tomorrow, who is legally exposed, and does your current permission model (scopes, spend caps, human approval steps) hold up under that assumption?