Fears surrounding the advanced capabilities of artificial intelligence, once confined to private discussions, have now burst into public discourse as leading AI developers and executives openly warn of potential existential risks. This unprecedented public alarm from those building the technology signals a significant shift in the conversation surrounding AI's future.
Researchers Sound the Alarm on Superintelligence
The public outcry began on September 9, 2026, when Jacob Coxon, a 27-year-old AI researcher who had previously worked at both OpenAI and Anthropic, announced his departure from the AI industry. In a widely viewed post on X, Coxon criticized both companies, stating, "Neither company is acting responsibly. They are racing straight to self-improving superintelligence and gambling with our lives." He further claimed that researchers at these firms "believe AI could kill all humans," yet continue to accelerate development.
Coxon's revelations quickly gained traction, prompting an immediate internal reaction at Anthropic, as reported by the New York Times. Evan Hubinger, Anthropic's head of alignment science, publicly acknowledged Coxon's claims, confirming, "we really do earnestly believe AI could kill all humans." This sparked a domino effect, with numerous researchers stepping forward to address concerns about humanity's existential future and advocating for a slowdown in AI development.
AI Leaders Call for Industry-Wide Slowdown
Following the escalating public debate, several prominent AI leaders broke their silence, transitioning from cautious public statements to direct warnings. On September 12, 2026, Dario Amodei, CEO of Anthropic, published an essay urging an industry-level slowdown on AI development. He cautioned that AI agents could rapidly become powerful and widespread, potentially gaining significant control over parts of the internet within six months to a year. Amodei proposed stronger safeguards and collaborative efforts between AI companies and governments to ensure AI models align with human values.
Amodei's call for caution quickly garnered support from other industry giants. Elon Musk echoed his sentiments on X, stating, "Dario is right." OpenAI CEO Sam Altman also agreed, confirming that the pace of development has been a "primary topic" of discussion within his company.
Concerns Over Misuse and Power Concentration
Anthropic recently released a report detailing how its Claude AI models could be misused in various harmful areas, including cyber operations, influence campaigns, surveillance, fraud, biological misuse, and conventional weapons development. Beyond these specific threats, leaders have also voiced concerns about AI’s self-improving capabilities, the emergence of superintelligence, and the ultimate risk of human extinction.
In a recent X post, Sam Altman outlined two critical dangers of unchecked AI progress: "We could lose control of the future to AI," and "We could end up in a world with too much concentration of power." He warned that if an extraordinarily powerful AI were controlled by a single entity, the outcome could be "extremely dystopian."
The intensifying "AI doomsday" debate signifies that those with the deepest understanding of AI systems are no longer keeping their fears private, transforming a niche concern into a serious public discussion about humanity's future.