Jensen Huang says AI apocalypse by 2030 has zero percent chance

By Splitfeed EditorialLast updated Sep 26, 2026, 21:18 UTC

On September 18, 2026, Nvidia CEO Jensen Huang delivered a keynote at GTC Washington that included new AI product announcements. The following day, a CBS Sunday Morning interview aired in which Huang stated there is a "0% chance" AI destroys the world by 2030, called warnings about AI dangers "unnecessary" and "irresponsible," and said Nvidia would develop AI "as fast as we can, irrespective of anyone else." He directly dismissed apocalyptic warnings made by Anthropic CEO Dario Amodei and OpenAI CEO Sam Altman, whose companies have each publicly argued for caution around advanced AI systems.

The remarks matter because of who is making them and why his position is structurally different from the safety advocates he criticized. Nvidia sells the chips that power virtually every large AI system being built today, meaning Huang profits directly from rapid development. Amodei and Altman, despite building commercial AI products themselves, have each testified before legislatures and published warnings about existential risk. Huang's dismissal of their concerns as irresponsible adds a sharp commercial dimension to what has been framed as a technical and ethical debate. The core dispute is whether publicly raising the possibility of catastrophic risk is a genuine public service or, as Huang argues, needless scaremongering that slows beneficial progress.

No specific reactions from Pro-AI, Anti-AI, or Middle Ground commentators were available at the time of writing. Typically, the Pro-AI camp would welcome Huang's framing, arguing that regulatory delay costs lives and economic progress, and that probability estimates of extinction by 2030 deserve scrutiny rather than deference. The Anti-AI camp would be expected to point to Huang's financial stake in the outcome as disqualifying his reassurances, and to argue that a "0% chance" claim is itself irresponsible given genuine uncertainty among researchers. The Middle Ground camp would likely acknowledge that near-term doom by 2030 is widely considered unlikely, while arguing that Huang's blanket dismissal of safety concerns forecloses legitimate debate about longer-term risks and governance.

The sharpest test of this disagreement will come if either Amodei or Altman responds directly to Huang's characterization of their warnings as irresponsible. Congressional hearings scheduled for late 2026 on AI regulation could also force these competing framings into a more formal arena, where Nvidia's lobbying position against new restrictions would sit alongside safety-focused testimony from the very lab heads Huang dismissed.