Home / Technology / AI Experts Warn: Humanity Faces Existential Threat
AI Experts Warn: Humanity Faces Existential Threat
14 Sep
Summary
- Superintelligence is likened to an uncontrollable alien mind.
- Several key AI researchers have resigned citing safety concerns.
- Leaders from OpenAI, Anthropic, and Grok acknowledge AI risks.
Concerns are escalating among AI researchers regarding the potential existential risks of superintelligence. Former Anthropic researcher Jacob Coxon has warned that the pursuit of AI more intelligent than humans is akin to summoning an alien species – an intelligence that could be fundamentally uncontrollable and incomprehensible.
Coxon's departure follows a pattern of insiders leaving major AI companies to voice their alarms. In February 2026, Mrinank Sharma, formerly of Anthropic, resigned due to worries about internal pressures compromising critical safety research for rapid deployment. This mirrors events at OpenAI, where Jan Leike and Ilya Sutskever departed over disagreements about the company's safety culture and research trajectory.
Daniel Kokotajlo, a governance and alignment researcher, also left OpenAI, citing a loss of confidence in leadership's ability to manage advanced AI systems responsibly. Other researchers have expressed similar concerns, with some refusing exit packages tied to non-disclosure agreements to maintain their freedom to speak out about safety issues.
