[Edaily Reporter Bang Sung Hoon ] Jensen Huang, CEO of NVIDIA, and Mark Zuckerberg, CEO of Meta, have rejected a proposal for the industry to collectively slow down the pace of artificial intelligence (AI) development. They argue that each company must take responsibility for addressing AI safety issues on its own.

According to the Financial Times (FT) on the 15th (local time), CEO Huang said at an event in San Francisco, California, alongside Salesforce CEO Marc Benioff, “Market forces are already at work. We don’t need new laws or new regulations.” CEO Zuckerberg also wrote on X (formerly Twitter) that same day, “Every lab has the responsibility and incentive to move at the pace necessary to train models safely, and the ability to take action on its own to do so.”

This is a response to the public proposal made last weekend by Dario Amodei, CEO of Anthropic; Sam Altman, CEO of OpenAI; and Elon Musk, CEO of SpaceX, who argued that research pace must be collectively slowed to properly address AI risks. While Google and Microsoft (MS) expressed their support, the two CEOs indicated they would not join the effort.

CEO Hwang emphasized, “If you’ve created a product or service but aren’t confident in its features, performance, or safety, don’t release it,” adding, “This is perfectly logical. You can simply adjust your own pace until you’re confident that what you’re releasing will be well-received by the market.”

He added, “Innovation and safety can certainly coexist,” noting, “Run as fast as you can, but if at any point you feel the company is spiraling out of control or the product might not be safe, just pause for a moment and correct the course.” He urged AI research labs not to fall into the “false dichotomy” of having to choose between the pace of innovation and the development of safe products. NVIDIA is also a major investor in OpenAI, Anthropic, and SpaceX.

CEO Zuckerberg echoed CEO Huang’s sentiments, stating, “Since we face massive legal liability if a model causes harm, there is a strong incentive to prevent that.” He also noted that Meta had delayed the launch of its AI agent “Muse” by several months, pointing out, “We didn’t demand that everyone else do the same before we did.”

However, CEO Zuckerberg believed it would be helpful if the ecosystem of external independent experts—known as “evaluators”—used by AI research labs to review models became larger and more diverse. He did not comment on Anthropic’s proposal to grant evaluators greater access.

Analysts say the two executives’ pushback highlights a deep-rooted conflict between “closed” model developers—such as OpenAI and Anthropic, which do not disclose their internal workings—and those who support “open” models that anyone can modify and use. CEO Hwang has consistently emphasized the importance of open models for the widespread adoption of AI, and Meta has also generally advocated for open models. Among open-model supporters, there are voices arguing that development regulations only benefit large research institutes that sell closed models.

CEO Hwang is aligned with U.S. President Donald Trump in opposing efforts to slow down the pace of AI development. He took a call from President Trump on stage at an event the previous day. When President Trump dismissed claims that AI poses a threat to humanity as a “scam” and criticized moves to slow down U.S. AI research and data center construction, CEO Hwang replied, “That’s right. We won’t let that happen.”

CEO Zuckerberg is also opposed to the idea of a self-regulatory body for the AI industry that Google, OpenAI, and Anthropic have been discussing. He reportedly conveyed his opposition to this plan during a phone call with President Trump this summer.