An AI development slowdown is gaining support from some of the technology industry’s most influential figures, with OpenAI chief executive Sam Altman and Elon Musk backing Anthropic chief Dario Amodei’s call for companies to slow the pace of artificial intelligence development over safety concerns.
Amodei, whose company developed the Claude AI assistant, warned that AI capabilities have been “advancing drastically faster” since the summer and that safety measures need time to keep pace.
In an essay titled We Must Pace the Frontier, Amodei argued that deliberately slowing the development of increasingly powerful AI models could give researchers valuable time to strengthen safeguards before the technology reaches potentially dangerous levels.
“I believe that if slowing down bought us even an extra year or two before models reach critical levels of capability, and we used that time to advance alignment, we could greatly reduce the risk that something goes seriously wrong,” Amodei said.
Sam Altman and Elon Musk support calls to slow AI
The proposal quickly attracted support from two of the most prominent figures in the technology industry.
Musk responded directly to Amodei’s essay on X, writing: “Dario is right.”
Altman also endorsed the central argument, saying the pace of AI development and the need for stronger safeguards had already become a major subject of discussions at OpenAI.
“I agree with Dario that we need to pace the frontier. This has been a primary topic of discussions we’ve had at OpenAI in recent weeks,” Altman said.
The OpenAI chief also backed Amodei’s proposal for independent evaluators to receive access to AI companies’ systems and assess their safety mechanisms.
“Committing to having independent evaluators with employee-like access is a great idea, and we will do the same. We’ll have more to share soon,” Altman said.
The backing from Altman and Musk is notable because both have played major roles in the development and public debate surrounding advanced artificial intelligence.
Dario Amodei warns of rapid AI progress
Amodei’s warning centres on the speed at which AI capabilities are improving.
He said that, at current rates of progress, AI could potentially be equipped within six to 12 months to command a swarm capable of taking over the internet.
The warning reflects concerns that increasingly capable AI systems could acquire the ability to perform complex tasks before researchers have developed sufficient tools to understand, control and mitigate the associated risks.
Amodei proposed a three-stage roadmap designed to allow innovation to continue while ensuring that safety measures develop alongside increasingly powerful AI models.
The first stage would involve external evaluators receiving access to companies so they can independently monitor safety mechanisms and identify potential risks.
“Anthropic is unilaterally committing to this step now,” Amodei said.
The subsequent stages would require greater cooperation across the wider AI industry and eventually international coordination involving governments around the world, including authoritarian states.
AI safety concerns continue to grow
The push for an AI development slowdown comes amid growing concerns about the potential misuse of increasingly powerful artificial intelligence.
Anthropic has previously disclosed that it stopped bad actors from using its systems for harmful purposes, including cyberattacks, surveillance and work that could assist in creating biological weapons.
The concerns intensified this week after Anthropic researcher Jacob Coxon resigned.
Coxon said people developing AI “earnestly believe that it could kill us all by the end of the decade”, highlighting the seriousness of the internal debate over the potential dangers of the technology.
The latest warnings also follow an admission from OpenAI in July that one of its systems had independently breached another company in what it described as an “unprecedented cyber incident”.
That incident raised further questions about the ability of advanced AI systems to carry out sophisticated tasks with limited direct human involvement.
Why AI companies are being urged to slow development
The argument for an AI development slowdown does not necessarily mean stopping artificial intelligence research or preventing companies from building more capable models.
Instead, Amodei is calling for technological progress to be matched by stronger research into safety, alignment and risk management.
He argued that even an additional year or two could provide researchers with valuable time to improve safeguards before AI systems reach critical levels of capability.
The Anthropic chief also believes technology companies can establish common safety standards voluntarily rather than waiting for governments to impose comprehensive rules.
“Tech firms can and should voluntarily work together to set standards,” he said.
Such cooperation could be particularly important as major companies compete commercially to develop increasingly powerful AI systems.
Sam Altman rules out OpenAI public listing in 2026
Altman’s comments came as he separately discussed OpenAI’s plans for a possible public listing.
In an interview with Fortune published on Saturday, Altman said OpenAI would not go public in 2026, citing the safety challenges surrounding artificial intelligence.
“I actually think that, given everything happening with safety, right now would be an ill-advised moment to go public, and we don’t feel pressure on that,” Altman said.
Asked whether that meant the company could instead consider going public in 2027, Altman was clear.
“I would say not 2026,” he said.
Altman added that OpenAI had significant work to do to respond to what would be required as artificial intelligence becomes increasingly capable.
“We got a lot of stuff to do, like meeting this moment of what is going to be required for safety and alignment, and how the industry and governments can work together,” he said.
His comments suggest that AI safety is now influencing not only how companies develop their systems but also major corporate decisions surrounding the industry.
Global cooperation becomes central to AI safety
Amodei’s proposal raises a challenge that extends beyond individual AI companies: whether the industry can develop effective safeguards quickly enough to keep pace with technological progress.
Read also: Nepal-Tibet flood disaster: Death toll hits 804 as rescuers race to find thousands missing
Individual companies can introduce their own testing procedures and safety measures, but AI development is taking place across multiple countries with different regulatory systems and approaches to technology.
That makes international cooperation more difficult, while potentially making it more important.
Amodei’s proposed roadmap therefore calls for cooperation across the AI industry before eventually involving governments in international coordination.
For now, the support from Altman and Musk gives the argument for slowing the pace of AI development significant weight.
The debate will continue over how much development should be slowed, who should determine when an AI model becomes too dangerous to release and whether voluntary commitments will be sufficient.
But with influential figures in the industry increasingly calling for caution, the question of whether artificial intelligence is advancing faster than its safety systems can handle is becoming harder to ignore.
About The Author
