UN’s IPCC Retreats from Climate Doomerism
There was a time when global warming arrived dressed for the Apocalypse. The poles would melt, the seas would rise, the forests would burn, crops would fail and humanity would discover that the four horsemen had traded their horses for sport utility vehicles. Every unusually hot day was a preview of the inferno, every cyclone a warning from the future, and every climate conference another gathering at which statesmen solemnly announced that there were only ten years, or perhaps five years, left to save the planet. Always the same figure as time moved on.
Something rather quieter has now happened. The climate apocalypse has not been cancelled, exactly. It has been moved from the main dining room to the lounge. The gentlemen are no longer standing on the deck shouting that the ship is sinking. They have settled into the leather chairs, loosened their ties and ordered a bottle of wine. Smoking is prohibited, unfortunately, but one suspects that if the regulations permitted it somebody might even produce a cigar.
The important development comes from the scenarios that will feed the next generation of global climate models. The Scenario Model Intercomparison Project has produced the new scenario architecture for CMIP7, whose modelling results will in turn become an important part of the scientific literature assessed in the IPCC's Seventh Assessment Report. The highest-emissions scenario is now considerably less extreme than the notorious SSP5-8.5 scenario used in the previous modelling generation. The authors of the CMIP7 scenario paper state quite plainly that the emissions levels represented by SSP5-8.5 have become implausible, citing falling renewable-energy costs, existing climate policies and actual emissions trends.
That deserves considerably more attention than it has received because SSP5-8.5, and before it RCP8.5, supplied much of the imaginative fuel for the climate-apocalypse era. These were scenarios in which fossil-fuel consumption and emissions climbed to extraordinary levels through the century. They were useful as high-end modelling experiments, but too often the distinction between a scenario and a prediction disappeared somewhere between the scientific paper and the newspaper headline. An extreme pathway became "what scientists say will happen," and impacts calculated under that pathway became photographs of our supposedly inevitable future.
The climate-modelling community is now saying that the upper boundary of what it regards as plausible has come down. The World Climate Research Programme's own explanation says that recent emissions have tracked much closer to the previous middle-of-the-road scenario than to the highest one, while renewable energy has become cheaper much faster than expected. Consequently, even its new scenario involving policy rollback and extensive fossil-fuel use produces lower emissions than RCP8.5 and SSP5-8.5.
The numerical change is not trivial. A recent analysis of the CMIP7 scenarios notes that the highest scenarios used across the previous three IPCC modelling generations produced roughly 4.6–4.9°C of warming by 2100 when evaluated on a comparable basis. The new CMIP7 High scenario comes in at about 3.3°C, with roughly half the cumulative carbon dioxide emissions of its predecessor. Meanwhile, the new Medium scenario, intended to represent something closer to a continuation of present policies, produces around 2.9°C by 2100.
The WCRP says the preliminary estimate for the High scenario is around 3.5°C by 2100 and explicitly warns that uncertainties could still produce warming above 4°C. Nor has the IPCC repudiated its fundamental position that human greenhouse-gas emissions are warming the climate and that additional emissions will produce additional warming.
Indeed, there is an irony at the other end of the graph. While the ceiling has come down, the floor has moved up. Scenario designers no longer regard a pathway that simply remains below 1.5°C as plausible. Even the most aggressive new pathway involves an overshoot before temperatures are brought back towards 1.5°C late in the century, requiring substantial carbon removal. The new picture is therefore compressed at both ends. The most extravagant fossil-fuel inferno has become implausible, while the immaculate 1.5°C future has also disappeared.
That is precisely why the development is culturally interesting. Climate politics flourished for years on the extremes. One side wanted catastrophe; the other wanted salvation. The political machinery was built around deadlines, tipping points and emergency language. The new modelling architecture looks more like the dreary territory in which most public policy actually lives: there will be warming, there will be costs, there will be adaptation, technologies will change, energy systems will evolve, policies will have effects, predictions will be revised and the world of 2100 will almost certainly surprise everybody presently attempting to model it.
The IPCC itself was quick to object when reports began saying that it had downgraded its predictions. Strictly speaking, the IPCC is right. It does not own these scenarios, does not run the climate models and has not yet assessed the CMIP7 results. Its job is to assess the scientific literature, and AR7 is still years away. In May 2026 it specifically issued a clarification making exactly that point. Yet the institutional distinction should not conceal the significance of what has occurred. The scenario machinery supplying an important part of the evidence base for the next assessment has changed because reality has overtaken assumptions embedded in the previous generation.
There is a lesson here that extends beyond climate science. Models are not crystal balls. Feed an extreme socioeconomic future into a climate model and the model can tell us something useful about the climatic consequences of that future. It cannot tell us that the socioeconomic future will occur. When journalists, activists or politicians silently convert "if emissions follow pathway X" into "scientists predict X," a conditional experiment becomes prophecy.
RCP8.5 demonstrated the problem magnificently. There were legitimate scientific reasons to run it. Researchers need boundary cases, just as engineers test structures beyond normal operating conditions. But nobody reports a bridge's extreme stress test by announcing that engineers predict every truck in the country will drive onto the bridge simultaneously next Tuesday. Climate communication was frequently less discriminating. Worst-case modelling acquired a peculiar journalistic privilege because catastrophe attracts attention, political urgency and money.
Now reality has performed an inconvenient experiment of its own. Renewable technologies became cheaper. Energy systems changed. Climate policies accumulated. Actual emissions failed to climb along the most extreme trajectory. The high-end scenario therefore had to be pulled downward. That is science behaving properly: assumptions are revised when the world refuses to cooperate with them. The embarrassment belongs not to scientists who make such revisions, but to anyone who previously sold contingent scenarios as inevitable futures.
There may also be a political reason why the temperature of the rhetoric seems to be falling. Western societies have discovered a newer apocalypse. Artificial intelligence can now perform the cultural function once occupied by climate catastrophe with remarkable efficiency. It has everything an emergency politics requires: incomprehensible technology, enormous corporations, frightening predictions, expert priesthoods, existential-risk calculations, international conferences and an inexhaustible supply of proposed legislation. Instead of asking whether Manhattan will be underwater, politicians can ask whether an artificial superintelligence will turn humanity into paperclips. The old climate countdown clock can be wheeled into the storeroom because AI has supplied a much more exciting one.
The legislative machinery is already warming up. Governments that spent years promising to manage the atmospheric temperature of the planet now propose to manage algorithms, foundation models, synthetic media, automated decisions and whatever comes after them. Climate change has the disadvantage of moving slowly enough for yesterday's predictions to be compared with today's observations. AI changes every few months. For bureaucratic emergency politics, that is an almost perfect renewable resource.
None of this means climate policy is disappearing. Governments have embedded enormous regulatory, financial and industrial structures around decarbonisation, and those structures will not vanish merely because the scenario range changes. Nor will climate conferences suddenly cease. Australia's Chris Bowen can presumably always find someone willing to discuss net zero with him. The bureaucracy will endure long after the banners have faded.
What may be disappearing is something more subtle: the monopoly of climate change upon Western apocalyptic imagination. The rhetoric that once presented global warming as the singular existential crisis of our civilisation increasingly has to compete with artificial intelligence, war, demographic decline, pandemics, energy insecurity and economic stagnation. Climate change is being pushed back towards being what it should always have been treated as: a serious scientific, technological and economic problem involving probabilities, trade-offs, adaptation and uncertainty rather than a secular Last Judgment.
The climate apocalypse has therefore not quite ended. It has matured. Having spent its youth marching through the streets predicting planetary death, it has reached middle age, discovered uncertainty ranges and taken a seat at the bar. The cigar will have to remain imaginary because somebody has already passed a regulation against it. But surely, after thirty years of saving the world, the gentleman may at least be permitted a glass of wine, to drown his sorrows!
https://www.gbnews.com/science/apocalypse-forecasts-climate-scientists
