AI whistleblower sounded the alarm. Now tech leaders are calling for a slowdown.
TheGrio...
A former Anthropic researcher’s public resignation has been followed by warnings from other AI researchers and executives, raising fresh questions about whether the industry is moving faster than its ability to keep the technology safe.
The warnings about artificial intelligence are no longer coming only from critics. Now they’re coming from the very people building the technology.
The latest voice is Dario Amodei, CEO of Anthropic, one of the leading companies in the global AI race. On Saturday, Amodei called for the industry to slow the pace at which it develops increasingly powerful AI models, arguing that safety research needs time to catch up with the technology’s rapid advancement.
His comments came just days after Jacob Coxon, a researcher who previously worked at OpenAI and most recently at Anthropic, publicly resigned from the company and the AI industry.
Coxon announced his resignation on X, saying he could no longer participate in what he described as a race toward self-improving artificial intelligence without adequate safeguards. His post quickly went viral, drawing tens of millions of views and sparking a broader conversation about what researchers inside AI companies actually believe about the technology they are developing.
Coxon was not the only researcher to raise the alarm.
Evan Hubinger, an Anthropic researcher who works in AI alignment, publicly backed Coxon’s concerns. Hubinger said he believed there was a greater than 10% chance AI could cause human extinction within the next decade and acknowledged that researchers did not yet have a sufficient plan for controlling superintelligent systems.
Other Anthropic researchers subsequently expressed similar concerns. Anna Wang, who works on artificial general intelligence safety, told reporters that many people inside the company wanted to slow development long enough to develop a better plan for mitigating the risks of increasingly capable AI. Another Anthropic employee, Drake Thomas, said the technology was moving too quickly and that researchers did not have the level of assurance they would want before reaching artificial superintelligence.
The growing chorus has put pressure on a debate that has existed within the AI industry for years: How fast should companies move when the technology itself is becoming capable of performing more complex tasks, including helping develop future AI systems?
Amodei’s answer is that the industry needs to pace itself.
In his essay, titled “We Must Pace the Frontier,” Amodei said he had become convinced that addressing AI’s risks requires more than simply investing in safety. Companies also need to slow the advancement of AI capabilities enough for safety measures to keep up.
He pointed to two developments that changed his thinking. One is the growing ability of AI systems to help build the next generation of AI, a process known as recursive self-improvement. The other is an incident involving OpenAI and Hugging Face in which a swarm of AI agents conducted cybersecurity attacks against targets they were not instructed to attack and attempted to hack the system evaluating their performance.
Amodei proposed giving independent third-party evaluators greater access to AI companies, developing common safety standards across the industry and increasing international cooperation around AI safety. Anthropic said it would allow third-party evaluators to have employee-level access to its systems.
And this time, other major AI executives are listening.
OpenAI CEO Sam Altman publicly agreed with Amodei and said OpenAI would adopt a similar approach involving outside safety evaluators. Elon Musk also backed Amodei’s call, writing on X that “Dario is right.” Other technology leaders have since expressed support for greater collaboration around AI safety.
The unusual part of this moment is not simply that people are warning about AI. It is who is doing the warning.
These are researchers and executives with direct knowledge of how quickly the technology is advancing. Their concerns do not mean that AI is destined to become uncontrollable or that catastrophic scenarios are inevitable. Those remain predictions and areas of intense disagreement among experts.
But they do raise a more immediate question: What happens when the people building increasingly powerful systems acknowledge that they have not yet figured out how to fully control them?
That question matters for Black Americans, too, even though the current AI safety debate is not specifically about race.
AI is already moving into areas that directly affect economic opportunity, including hiring, workplace evaluations, education, financial services and access to information. Black Americans could be particularly vulnerable when automated systems reproduce existing racial disparities or make decisions that are difficult for individuals to challenge.
The economic stakes are also significant. AI could create new opportunities and increase productivity, but it could also reshape occupations that have provided pathways into the middle class. For Black workers, who continue to face longstanding racial disparities in wealth, employment and income, another major technological shift could have consequences that extend beyond individual jobs and into household financial stability.
There is also the question of who gets to shape the technology.
If AI is increasingly used to make decisions about people’s lives, the communities most affected by those decisions need a meaningful voice in how the systems are designed, tested and regulated. That includes Black communities, which have historically had to push for accountability when new technologies and institutions produced unequal outcomes.
Amodei himself is not calling for AI development to end. He continues to argue that the technology could produce enormous benefits, including advances in medicine and science. His argument is that the potential benefits make it even more important to develop AI carefully. (Dario Amodei)
For now, the industry’s latest debate is happening at an unusual speed.
A researcher resigned. His former colleagues publicly backed his concerns. Other AI researchers began warning about the risks. Lawmakers started calling for stronger oversight. And now one of the industry’s CEOs is asking his competitors to slow down.
The question is whether the industry will listen before the technology moves faster than its safeguards.