Jensen Huang at Dreamforce: AI Needs No New Laws

The 2026 Dreamforce stage offered a rare back-to-back contrast in how leading AI firms speak about risk. Anthropic's Dario Amodei, whose company is a critical partner helping usher Salesforce into the AI era, used his appearance to push for industry-wide standards rather than competitor attacks after what the piece describes as OpenAI's Hugging Face incident. Minutes later, Nvidia's Jensen Huang downplayed those concerns and argued companies can simply police themselves, telling the room they should "run as fast as you can" and pause only if their own products become unsafe.

The debate extended beyond the keynote stage. AI researchers Sayash Kapoor and Arvind Narayanan published one of their longest essays, arguing the real crisis is more narrowly about cybersecurity and organizational maturity, with Kapoor calling for AI labs to be made liable for harms. Sydney Von Arx of the safety nonprofit Nightingale, conversely, said the same incidents show many current AIs are "egregiously misaligned."

What the episode leaves unresolved is whether the companies pushing frontier AI will actually exercise the self-restraint Huang described. In the same week they warned about slowing down, the article notes, the same CEOs were at Dreamforce pitching their latest products and looking to keep the money flowing. How strictly planned AI regulation in Congress advances is likely to shape whether the industry's self-policing argument holds up.