Story perspectives

Jensen Huang Says Unsafe AI Risks Civil, Criminal Liability

By Drooid · Sep 25, 2026, 6:06 AM UTC · How we work

Summary

Story summary

Jensen Huang said AI does not pose an existential threat and called risk warnings a distraction.
He warned that releasing unsafe AI could expose firms to civil and criminal liability.
He predicted Nvidia will double its chip sales next year.
He said transitioning to carbon-free energy will take several years, requiring continued fossil fuel use.
He noted AI development will raise Nvidia’s total AI industry investment to about $100 bn.

History

Historical Context

In February 2019 OpenAI announced GPT-2 but withheld the 1.5-billion-parameter model, citing misuse risks; the limited version was released later that year while the full model appeared on 5 November 2019. This staged rollout echoed earlier safety debates, such as the Allen Institute’s fake-news detector, and influenced later AI safety policies at OpenAI.

1 of 2

Story summary

Jensen Huang said AI does not pose an existential threat and called risk warnings a distraction.
He warned that releasing unsafe AI could expose firms to civil and criminal liability.
He predicted Nvidia will double its chip sales next year.
He said transitioning to carbon-free energy will take several years, requiring continued fossil fuel use.
He noted AI development will raise Nvidia’s total AI industry investment to about $100 bn.

1 / 2