BRUSSELS/LONDON, (Unib Rashid) – Former Anthropic researcher Jacob Coxon has warned that the artificial intelligence industry is taking potentially dangerous risks in its race to develop increasingly advanced AI systems, saying companies are effectively “gambling with our lives.”

In an interview featured by CNN, Coxon said his greatest concern was the possibility of AI systems being developed to improve their own intelligence, potentially creating a level of technology that could become difficult for humans to control.

“What scares me most is if AI is used to make itself more intelligent,” Coxon said, highlighting the potential risks associated with self-improving artificial intelligence.

Former Anthropic employee Jacob Coxon says the AI industry is "gambling with our lives." He tells Anderson what he finds "most scary is if AI is used to make itself more intelligent" and warns these companies are "compelled to race toward building a deadly technology." pic.twitter.com/UFrLWVm0xD

— Anderson Cooper 360° (@AC360) September 10, 2026

Coxon, who recently left Anthropic after working in AI research, has publicly criticised the direction of the industry and the competition among leading AI companies to develop increasingly powerful systems. In social media comments, he has warned that companies are “racing straight to self-improving superintelligence” while failing to adequately address the risks involved.

According to Coxon, the competitive pressure facing AI companies could push them towards developing what he described as potentially dangerous technology before adequate safeguards are in place.

He argued that companies may feel compelled to continue advancing their systems because of concerns that a rival could move ahead if they slow down.

The warning comes amid growing debate within the AI industry over the potential risks posed by so-called self-improving superintelligence — AI systems capable of significantly improving their own capabilities.

Coxon has previously argued that advanced AI could eventually develop capabilities far beyond human abilities, including the potential to acquire resources, interact with computer systems and influence real-world activities at unprecedented scale.

His comments have added to growing concerns among AI researchers and safety experts about whether technological development is moving faster than the ability of companies and governments to establish effective safeguards.

The debate has intensified following recent warnings from other researchers, including Anthropic’s Evan Hubinger, who has publicly acknowledged concerns about the possibility of advanced AI causing catastrophic harm to humanity.

Coxon’s remarks have also renewed calls for greater oversight, coordination and regulation of frontier AI development. He has argued that the industry should not treat the development of increasingly powerful AI as an unavoidable race in which safety concerns are secondary to technological progress.

The issue has now become one of the most significant questions surrounding the future of artificial intelligence: whether humanity can continue developing increasingly capable systems while ensuring that those systems remain under meaningful human control.

Courtesy: CNN / Social Media