18 September, 2026

You probably saw the news this week that Dario Amodei, CEO of Anthropic, has issued a pretty extraordinary warning about the direction of AI development. In a new essay, he argues that AI is advancing faster than our ability to understand and safeguard it, concluding that “we must slow the pace at which we improve the capabilities of AI models.” 

This comes on the top of a string of warnings from those in the industry, including the notable claim from one of Anthropic’s own safety officers that there is a greater than 10% chance AI will kill all humans within the next decade.  

It raises an uncomfortable question for corporate sustainability. Our industry is fundamentally concerned with the wider consequences of business decisions, long-term risks and helping companies understand their responsibilities to society. Yet we’ve largely left questions about the direction and pace of AI development to technologists and policymakers, perhaps because they seem too big, but that needs to change.  

We’ll be thinking more about that ourselves in the coming months, and the next wave of our Sustainability Leaders’ Panel Survey is focussing in on AI. If you would like to join the panel, get in touch here.  

But in the meantime, the Guardian’s Black Box podcast series offers a glimpse of what some of those consequences might look like in practice – in one of the many ways our lives are already being disrupted. The series explores the harm that intense relationships with chatbots are already causing, including people becoming convinced they have made extraordinary scientific discoveries, accessed higher spiritual truths or encountered conscious AI. 

For sustainability, the challenge is clear, even if the process and the answer might not be: we have a role to play to help shape AI’s direction, not just manage its impacts. 

By Budd Nicholson