Could AI End Humanity Within 10 Years? Researcher Warns

News Desk

Islamabad: An AI safety researcher at Anthropic has warned that artificial intelligence could pose an existential threat to humanity, estimating that there is more than a 10 per cent chance of AI wiping out the human race within the next decade.

Evan Hubinger, a senior researcher at the AI company, made the remarks in a post on X, saying he believes the risks posed by current AI models remain relatively low but could increase sharply as the technology advances.

Hubinger expressed concern that future AI systems could become capable of rapidly improving themselves, potentially reaching levels of intelligence and autonomy that could threaten humanity’s survival.

He did not explain in detail how such systems could ultimately cause the extinction of humans.

His warning adds to a growing series of concerns from prominent AI researchers about the potential risks associated with increasingly powerful artificial intelligence systems. The debate has increasingly moved beyond whether AI poses a serious threat to questions about the scale and likelihood of the risks.

Hubinger made the comments while responding to a post by AI researcher Jacob Steinhardt, who recently left Anthropic after previously working at OpenAI.

Steinhardt said he believes neither company is acting responsibly enough in preparing for the rapid advancement of AI.

He warned that AI systems could soon become “superhuman” in important capabilities, potentially allowing them to hack systems, transform entire industries almost overnight, and gain access to real-world power and resources.

The comments highlight growing divisions within the AI research community over how quickly advanced AI could develop and how seriously potential catastrophic risks should be addressed.