For years, the discourse around global existential risk centred almost entirely on climate change — a slow-burning catastrophe driven by human activity that demanded urgent, coordinated action. Now, artificial intelligence has stormed into the conversation with a velocity that has left policymakers, technologists, and the public grappling with a second, equally profound challenge. Both threats are human-made. Both are terrifying. And the world appears to be scrambling to catch up with the second.
A Planet Reckoning With Two Giants
Climate change has dominated the global agenda for well over a decade. Summits, treaties, carbon targets, and protest movements have all revolved around curbing greenhouse gas emissions and adapting to a warming planet. The science is unequivocal. The consequences — rising sea levels, extreme weather, biodiversity collapse — are already reshaping communities across the globe.
Yet just as the world appeared to be settling into a grim but familiar rhythm of climate negotiations, a new frontier emerged. Artificial intelligence, particularly large-scale generative systems, has moved from the margins of academic research into the fabric of everyday life within the space of mere years. The pace of development has outstripped the capacity of institutions to understand, regulate, or even debate its implications.
It is worth pausing on that disparity. Humanity had generations to contend with the mechanics of a changing climate. Against AI, the timeline feels almost compressible.
Silicon Valley Finds Itself in Uncharted Territory
The technology sector has always thrived on disruption, but this moment carries a different weight. Leaders in Silicon Valley — the very architects of the AI revolution — are now openly voicing concerns about their own creations. Statements from chief executives and researchers about the “existential risk” posed by advanced systems have shifted from fringe speculation to mainstream discussion.
This is a striking reversal. For much of the past decade, the tech industry’s narrative was one of boundless optimism: AI would democratise education, accelerate medical research, and unlock solutions to humanity’s thorniest problems. That optimism has not vanished, but it now sits alongside a far more uneasy awareness. The tools being built today are capable of generating convincing text, imagery, and code. Tomorrow’s systems may operate in ways their own creators struggle to fully explain.
The tension at the heart of Silicon Valley is unmistakable. Companies are investing billions in AI infrastructure while simultaneously calling for oversight. Engineers who built the systems now sit on advisory panels urging caution. It is a peculiar kind of reckoning — one born not from external pressure alone, but from an internal recognition that the genie cannot simply be guided back into the bottle.
The Regulatory Gap
Governments around the world are racing to formulate responses, but the legislative machinery of democratic states was not designed for this speed. The European Union has taken the lead with its AI programme, aiming to categorise systems by risk level and impose obligations on developers of high-impact models. The United States has leaned heavily on executive guidance and voluntary commitments from technology firms. China has moved with its own regulatory framework, focused on algorithmic governance and data security.
None of these approaches can be considered fully mature. The fundamental challenge is that regulation, by its nature, lags behind innovation. Lawmakers require time to study a technology, consult experts, draft legislation, and build consensus. AI development operates on a fundamentally different clock. Each successive model can arrive before the previous one has been thoroughly examined.
What makes this particularly acute is the question of coordination. Climate change, for all its political difficulties, is a threat with fairly well-understood mechanisms. International climate agreements — however imperfect — provide a framework for collective action. No equivalent global architecture exists for artificial intelligence. The conversations happening in Geneva, Brussels, Washington, and Beijing remain largely parallel rather than convergent.
Where the Two Crises Converge
There is a dimension of this story that often goes unexamined: the intersection of AI and climate change. Advanced machine learning systems require enormous computing power, and the data centres that house them consume vast quantities of energy. A single training run for a frontier model can emit hundreds of tonnes of carbon dioxide. The tech industry’s hunger for computational resources is, in part, fuelling the very crisis it claims to want to solve.
At the same time, AI is being deployed as a tool in the fight against climate breakdown. Researchers use machine learning to optimise energy grids, model weather patterns with unprecedented accuracy, and develop new materials for carbon capture. The technology is neither purely part of the problem nor purely part of the solution. It occupies a complicated middle ground.
This duality matters enormously. If the world is to navigate both challenges simultaneously, it cannot afford to treat them as separate conversations. They are deeply entwined, and policy responses must reflect that reality.
The Human Element
Beyond the geopolitics and the technology itself lies a more fundamental question about how societies respond to threats they do not fully understand. Climate change could be photographed, measured, and mapped. Its effects could be pointed to with a degree of specificity — this flood, this drought, this wildfire. Artificial intelligence, by contrast, is abstract, layered, and often opaque even to those who build it.
That abstraction makes public engagement harder. People can rally behind a cause when they can see its effects. Asking millions of citizens to care about the alignment of a neural network operating at a scale they cannot visualise is a different kind of challenge entirely.
Yet the stakes are, in some respects, comparable. Both crises demand a willingness to sacrifice short-term gain for long-term survival. Both require co-operation across borders and ideologies. And both expose the limits of institutions that were not designed for threats of this magnitude.
Why it Matters
The emergence of artificial intelligence as a co-equal threat alongside climate change is not a footnote in the story of the twenty-first century — it is one of its central chapters. How societies choose to respond will shape economies, power structures, and daily life for generations. The window in which thoughtful, deliberate action can still steer outcomes is narrowing, and the decisions made in boardrooms and legislative chambers over the coming years will carry consequences that far outlast the political cycles in which they were taken. This is not a moment for caution alone; it is a moment for clarity.