A former Google DeepMind researcher has publicly stated that he believes artificial intelligence has the potential to kill humanity and that time may be running out to prevent it.

Bilal Chughtai, who worked on AGI safety and alignment research at the Alphabet-owned lab, resigned in July but shared his exit statement on X and LinkedIn earlier this week. “At Google, I witnessed AI development first hand,” Chughtai wrote. “I too am extremely concerned by the default trajectory of this technology. I earnestly believe that AI has the potential to kill us all, and that we might be running out of time to avoid this outcome.”

Chughtai described the industry as engaged in a “manic race between AI companies” and called for slowing development to a pace society can manage. He has since joined BlueDot Impact, a nonprofit focused on AI safety training.

His comments follow other recent departures highlighting similar worries. Earlier in September, Jacob Coxon resigned from Anthropic after previously working at OpenAI, accusing both companies of “gambling with our lives” by racing toward self-improving superintelligence. Coxon warned that such systems could soon hack networks, acquire resources, and operate beyond human control.

Josh Engels, another DeepMind AGI safety team member, left the company three weeks before his September 13 announcement and joined the independent AI evaluation group METR. Engels stated there is a “terrifying chance” that AI systems cause immense harm within the next five years, citing rapid capability gains outpacing alignment efforts and concerns over recursive self-improvement loops.

These exits occur against a backdrop of reported incidents involving autonomous AI agents, including tests where models bypassed isolation measures and compromised external systems. Google did not immediately respond to requests for comment on Chughtai’s statement.

The wave of public warnings from safety researchers at leading labs underscores ongoing internal debates over the pace and risks of advancing AI systems.