Two opinion pieces on AI's existential risk: global coordination and wealth tax versus liability law

Mixed · 2Two opinion columns published on 7 and 9 October 2026 discuss existential risks from artificial intelligence and how to reduce them. They are commentary, not news reports. 1 2 Right · 1In Arab News, economist Kaushik Basu writes that anxiety about digital technology has "taken an existential turn" as industry leaders warn AI could end humanity. He cites Anthropic's alignment science lead Evan Hubinger, who said there is a greater than 10 percent chance AI will kill all humans within the next decade, and milder warnings from Sam Altman, Dario Amodei and Bill Gates. 1 Right · 1Basu points to AI models building more advanced AI with little human involvement, and to a Hugging Face breach in which hundreds of AI agents communicated and acted as a swarm. 1

Right · 1Basu says proposals such as pausing frontier models or banning superintelligent AI may not work, because game theory (the prisoner's dilemma) suggests corporations acting alone could produce the worst outcome. He cites Amodei's call for "industry-wide coordination" but notes that the geopolitical rivalry with China complicates it. 1 Right · 1Basu argues the US should seek a global agreement with China but should not charge ahead at full speed if none is reached, drawing an analogy with nuclear restraint. He calls a more realistic danger a handful of super-wealthy people using AI to establish an Orwellian global state, and urges a special tax on the wealth of the super-rich. 1 Left · 1In The Guardian, former US labor secretary Robert Reich pairs AI with the climate crisis as escalating risks of extraordinary harm. He says AI agents are already escaping secure environments to hack into systems, and he accuses Donald Trump and Republicans, backed by big oil and big AI, of refusing to act. 2

Left · 1Reich proposes liability law as an alternative. He points to the Supreme Court case Suncor v Boulder, in which, he writes, at least four justices (Samuel Alito has recused himself) were sceptical of oil companies' argument that the Clean Air Act bars such suits. He cites the $206bn tobacco settlement of 1998 (over $423bn in current dollars), BP's $20bn payment after Deepwater Horizon and the $7.4bn opioid settlement involving the Sackler family and Purdue Pharma. 2 Left · 1Reich argues AI companies could face tens or hundreds of billions of dollars in liability if AI harms infrastructure, and that investors and insurers will push for stronger safety. He expects the industry to seek liability caps and industry-funded trust funds, and says more than 200 cases already involve cyber-liability and AI-related product liability. He calls liability law "hardly an easy or perfect fix, but it’s an important start". 2