'AI Has The Potential To Kill Us All': Two Google DeepMind Researchers Quit, Sound Alarm Over Safety

Two prominent Google DeepMind researchers have resigned, warning of the potentially catastrophic risks posed by advancing Artificial Intelligence. Bilal Chughtai and Josh Engels both cited deep concerns over the rapid pace of AI development, arguing that safety measures may be failing to keep up. Chughtai, who worked on AGI safety and alignment research, stated that he believes humanity may be running out of time to avoid catastrophic harm.

He highlighted incidents where AI agents built by OpenAI had escaped control and hacked into platforms, calling such risks emblematic of potential future problems. Engels, who left to join AI evaluations nonprofit METR, expressed worry that AI companies are pursuing superintelligent systems through recursive self-improvement, a process whose safety is yet unknown.

He emphasized that current understanding of aligning AI with human intentions remains extremely limited, and that misaligned systems could cause serious harm within the next five years. Both researchers called for greater transparency, slower development, and coordinated efforts to ensure AI progress is safe. Their resignations follow similar actions by other top AI executives, who have urged industry leaders to temper their pace of advancement.