The United Nations has unveiled a new global framework aimed at forcing the world’s largest artificial intelligence companies to disclose the full environmental impact of their rapidly expanding data‑centre operations by the end of the decade. The initiative, described by the UN as an AI “green charter”, was launched by Secretary‑General António Guterres during London Climate Action Week on 23 June 2026.
The announcement comes amid mounting concern over the soaring energy and water demands of AI infrastructure. According to the UN, data centres powering advanced AI systems could, by 2030, consume more electricity than all but five countries and use enough water annually to meet the basic needs of more than a billion people in sub‑Saharan Africa. These projections have intensified pressure on governments and regulators to scrutinise the environmental footprint of the technology sector.
Under the new charter, AI developers and cloud‑computing providers will be required to publish detailed assessments of their carbon emissions, energy consumption, water usage and land impact. The UN is also urging companies to commit to operating entirely on renewable energy by 2030, arguing that transparency alone is insufficient without a clear pathway to decarbonisation.
Guterres warned that the unchecked expansion of AI infrastructure risks undermining global climate goals unless the industry adopts strict environmental reporting standards. He emphasised that the benefits of AI cannot come at the expense of the planet, calling the charter a necessary step to ensure that technological progress aligns with the UN’s broader sustainability agenda.
The initiative has been welcomed by climate researchers, who have long argued that the environmental cost of AI has been obscured by a lack of mandatory disclosure. Industry reaction has been mixed, with some major firms signalling support for clearer reporting rules, while others have raised concerns about the commercial sensitivity of operational data.
