Address concerns over potential AI threats

AI has potential to streamline every aspect of healthcare.[Courtesy]

Within a few years, Artificial Intelligence (AI) has moved from a novel discovery to a force actively reshaping how societies work, communicate, and make decisions.

AI’s benefits are real and tangible. Systems accelerate scientific discoveries, enhance medical diagnostics, and automate tasks once demanding immense effort. Handled responsibly, AI could drive genuine breakthroughs in productivity.  However, despite the technolgy’s obvious advantages, an urgent debate is unfolding among scientists and politicians on how far it should be allowed to go before serious safeguards are put in place.

Driven by corporate rivalries, private entities race forward with minimal public oversight. Two distinct categories of risk manifest. The first is immediate: Widespread job displacement, rampant misinformation, privacy erosion, cyberthreats, and the environmental costs of massive data centres. These demand firm government regulation today.

The second threat is existential: The fear that artificial superintelligence could entirely surpass human intelligence. Creating systems capable of outthinking humans across every domain could pose a serious threat to our safety and even existence, according to some scientists.

Early this month, Jacob Coxon, an AI researcher who worked atAnthropic announced tendered resignation, claiming that AI labs were “gambling with our lives,”. He said that the kill people building AI believed that it could kill all by the end of the decade.  If you think those were the idle musings of a young scientist,  Anthropic’s Alignment Science Lead Evan Hubinger has also warned of a greater than 10 per cent chance that AI could kill all humans within a decade.

Former UK Defence Secretary Des Browne and Berkeley professor Stuart Russell have similarly compared superintelligent AI to nuclear weapons, warning of a “Chornobyl-sized catastrophe” where humanity irreversibly loses control.

However, researchers like Dr Andrew Rogoyski suggest advanced AI is nowhere near human versatility and may face a “great disappointment” due to high costs, while Oxford’s Prof Sandra Wachter argues that existential “Terminator scenarios” distract from present-day harms. But that is not consoling enough.

The blood-curdling warnings issued by the scientists who fear that a doomsday scenario is approaching fast should make the world to pause and think about the direction we are headed. 

World governments should demand an immediate halt to the development of self-improving AI models until these concerns are addressed. Leaving the terms of AI development to private boardrooms racing for market advantage is unacceptable.

We need binding safety standards, independent audits, and international coordination to ensure AI remains a productive tool serving humanity rather than an uncontrolled threat that would in the end annihilate mankind. Our window to pause and regulate AI is right now, while humans still have control. Waiting for unambiguous proof of disaster before establishing safeguards would be reckless against the slow pace of formulating laws.