A British AI researcher with stints at two of the world’s most high-profile AI development firms, OpenAI and Anthropic, has sent shockwaves across the global tech industry after blowing the whistle on widespread internal fear over unregulated AI progress. Jacob Coxon, who recently left his post at Anthropic, took to X, the platform formerly known as Twitter, to publicly air his alarms that both leading companies are recklessly “gambling with humanity’s future” in their competitive rush to roll out increasingly powerful generative AI models.
In one viral accompanying post, Coxon made a staggering claim: the engineers and researchers actually building cutting-edge AI systems widely believe the technology could kill all human life by the end of this decade. The post racked up more than 100 million views in just 18 hours, and had crossed nearly 170 million views by the time this report was published, turning a long-simmering debate over AI safety into a global public conversation overnight.
The sudden viral attention has reignited long-simmering demands for tighter AI regulation and an industry-wide slowdown in advanced model development, triggering major ripple effects across the tech sector. While calls for greater caution around AI risk are not new, this unprecedented surge in public attention has forced top industry leaders to address the concerns head-on, as global audiences now pay closer attention to AI development than ever before.
Over the weekend, Anthropic CEO Dario Amodei penned an essay titled *We Must Pace the Frontier*, where he echoed the growing anxiety, noting that AI progress has accelerated “drastically faster” than expected since this summer. According to reporting from The Independent UK, Amodei warned that at the current rate of advancement, an unregulated AI swarm capable of hijacking the entire global internet could emerge within as little as 6 to 12 months. His warning follows multiple high-profile AI-related hacking incidents, the first major reported case occurring in July that affected both OpenAI and leading AI community platform Hugging Face.
In his essay, Amodei laid out a three-part framework to slow AI development responsibly while prioritizing safety. “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 wrote.
Remarkably, the call for a paced approach has received public backing from some of the biggest names in the industry. X and SpaceXAI owner Elon Musk replied simply to Amodei’s X post with “Dario is right”. OpenAI CEO Sam Altman also publicly endorsed the position, writing on the social platform: “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.”
A core part of Amodei’s proposal calls for independent safety evaluators to be granted employee-level access to AI development workspaces, to monitor safety protocols and flag risks publicly. For his part, whistleblower Coxon told the BBC that while he strongly supports an industry slowdown, any effective pause will require global coordination—specifically with major AI power China—to prevent a destabilizing international AI arms race that puts all nations at risk.