An OpenAI engineer publicly acknowledged feeling an “existential threat” from artificial intelligence, saying its disruptive impact on jobs, society and human relevance is a matter of “when, not if.” Hieu
Pham, a Member of Technical Staff at OpenAI, shared his concerns in a post on X (formerly Twitter), reflecting growing unease even among those building the world’s most advanced AI systems. Hieu Pham has previously worked at xAI, Augment Code and Google Brain.
“Today, I finally feel the existential threat that AI is posing,” Hieu Pham wrote, adding, “When AI becomes overly good and disrupts everything, what will be left for humans to do? And it’s when, not if.”
Hieu Pham’s remarks come at a time when AI systems are rapidly advancing across coding, writing, research and complex reasoning tasks. As companies such as OpenAI, Google, xAI and Anthropic push toward increasingly powerful models, concerns over job displacement, social disruption and loss of human control are intensifying across the tech industry.
Anthropic Safety Lead’s Exit Sparks Debate
Hieu Pham’s comments follow a similar warning issued earlier this week by Mrinank Sharma, who led Anthropic’s Safeguards Research Team before resigning on February 9. Mrinank Sharma had joined Anthropic in 2023 and worked on safety-focused AI research.
In a public letter announcing his departure, Mrinank Sharma wrote, “The world is in peril. And not just from AI, or bioweapons, but from a whole series of interconnected crises unfolding in this very moment.” He cautioned that humanity may be nearing a critical threshold where its wisdom must grow as fast as its technological power, warning of severe consequences if it does not.
Mrinank Sharma also hinted at tensions within AI companies, suggesting that corporate incentives and real-world pressures often make it difficult to prioritise safety and ethics, despite public commitments to responsible AI development.
Geoffrey Hinton, often referred to as the “godfather of AI” for his pioneering work in deep learning, has repeatedly cautioned that advanced AI systems could become uncontrollable, especially if they surpass human intelligence without sharing human goals. He has warned that once such systems emerge, “the idea that you could just turn it off won’t work” and has expressed regret over the speed at which AI capabilities are advancing.
