At Salesforce’s Dreamforce conference in San Francisco on 15 September 2026, Anthropic chief executive Dario Amodei called for an industry‑wide slowdown of AI development and outlined three concrete actions, while Nvidia chief executive Jensen Huang responded minutes later, urging developers to “run as fast as you can.” The exchange put two of the sector’s most powerful voices at odds on the pace of progress.
Dreamforce showdown
The Guardian reported that Amodei used an automotive analogy, saying a safety incident at a competing car maker is a moment for all manufacturers to pause and review practices. He framed the AI safety debate as a similar inflection point, arguing that “when a competing car company has a safety incident… it is a moment for all car companies to stop and review their own practices.”
Shortly after, Huang countered, “Run as fast as you can,” positioning Nvidia’s rapid‑innovation model as the antidote to what he described as an over‑cautious industry stance. Both remarks were made on the main stage of Dreamforce, a high‑profile tech event that draws global attention.
Amodei’s three‑step proposal
Amodei’s slowdown call was anchored in three specific actions:
- Embedding third‑party evaluators inside AI firms to provide employee‑level access for safety verification.
- Coordinating safety standards among democratic countries, creating a shared regulatory baseline.
- Eventually expanding coordination to a larger global framework.
The Guardian excerpt confirms the wording: “Amodei is calling for three courses of action: embedding third‑party evaluators inside AI companies; coordinating safety standards among Democratic countries; and, eventually, larger global coordination.” The proposals aim to create external checks without stifling innovation, according to Amodei.
Huang’s acceleration argument and Nvidia’s scale
Huang’s brief rebuttal underscored Nvidia’s belief that speed is essential to maintaining competitive advantage. The argument aligns with Nvidia’s financial footing: the company reported $177.837 billion in revenue for the six‑month period ending 26 July 2026 (Form 10‑Q, filed 26 August 2026), and net income of $118.010 billion for the same period. Total assets stood at $320.272 billion, with shareholders’ equity of $228.984 billion. Shares outstanding were 24.304 billion as of 25 January 2026 (Form 10‑K, filed 25 February 2026).
These figures illustrate Nvidia’s deep cash reserves and market‑cap leverage, giving the firm the bandwidth to argue for rapid development without immediate regulatory constraints.
| Metric | Value | Period End | Source |
|---|---|---|---|
| Revenue | 177,837,000,000 | 26 Jul 2026 | Form 10‑Q, filed 26 Aug 2026 |
| Net income | 118,010,000,000 | 26 Jul 2026 | Form 10‑Q, filed 26 Aug 2026 |
| Total assets | 320,272,000,000 | 26 Jul 2026 | Form 10‑Q, filed 26 Aug 2026 |
| Shareholders’ equity | 228,984,000,000 | 26 Jul 2026 | Form 10‑Q, filed 26 Aug 2026 |
| Shares outstanding | 24,304,000,000 | 25 Jan 2026 | Form 10‑K, filed 25 Feb 2026 |
Sector implications
The clash highlights a strategic fault line for investors. Companies that align with Amodei’s safety‑first approach may face tighter regulatory scrutiny but could benefit from early compliance advantages. Conversely, firms echoing Huang’s acceleration mantra may capture market share faster, leveraging Nvidia’s robust balance sheet to fund next‑generation GPUs and AI infrastructure.
Anthropic, founded in 2021 and employing roughly 2,500 staff, is still early in its growth trajectory. Its call for third‑party evaluators could set a precedent that reshapes how venture capital and corporate partners assess risk. Nvidia, with over 11,500 employees and a long‑standing presence in the semiconductor industry, is positioned to continue driving hardware demand regardless of policy shifts.
What remains unknown
The Guardian article does not disclose how many AI firms have already adopted third‑party evaluator access, nor does it quantify the potential cost of implementing such oversight. Likewise, Huang’s “run as fast as you can” stance is not backed by specific timelines or investment commitments, leaving investors to infer the pace from Nvidia’s existing financial strength.
Both CEOs left open the question of how democratic‑country coordination would be operationalized. No timeline was provided for moving from regional standards to a broader global framework, and no regulatory body was named as the lead coordinator.
As the AI sector grapples with safety concerns and market pressure, the Dreamforce debate may serve as a bellwether for policy direction and capital allocation in the months ahead.