Nvidia's Jensen Huang argues AI safety should be left to individual companies, not government regulation
2026-09-15
Per TechCrunch, Nvidia CEO Jensen Huang said AI is 'just hardware and software' rather than an 'alien mind,' and argued safety can and should be engineered by each AI product maker rather than mandated through government regulation.
Significance 3: A public, on-the-record policy position from the industry's dominant compute supplier at a moment when agentic-safety incidents are accumulating elsewhere, though Huang and Nvidia are not among the tracked labs and this is a single press account of one statement.
Implications · machine-drafted, not owner judgment
As a stated position from the infra layer that both funds and depends on unconstrained AI buildout, Huang's self-regulation argument is worth reading as a directly self-interested claim, per this tracker's standard discipline, rather than a neutral technical judgment — and it lands right as this month's evidence base (rogue-agent incidents, undisclosed attacks) makes the case for external oversight more concrete, not less.
- Whether any tracked lab (OpenAI, Anthropic, Google DeepMind) publicly agrees or disagrees with Huang's framing
- Movement on pending U.S. federal AI safety legislation in the following weeks
Sources
- press We don't need AI regulation — leave safety to us, Nvidia's Jensen Huang says retrieved 2026-09-16