As the UN demands transparency from AI companies, HR executives face questions about the environmental footprint of their technology investments
The United Nations (UN) is pressing artificial intelligence (AI) companies to disclose the true environmental cost of their operations, bringing more questions for organizations that are accelerating AI adoption.
On June 23, UN Secretary-General António Guterres launched the UN's AI Environmental Transparency Initiative during London Climate Action Week, calling on AI companies to publicly report their water consumption, carbon emissions, land use impacts, and energy requirements, the Associated Press reported. The initiative adds a new dimension to the conversation HR executives are having: what does it actually cost — in every sense — to transform an organization using AI?
The timing matters. Many Canadian employers are in the midst of significant technology procurement decisions, and environmental, social, and governance (ESG) considerations are increasingly embedded in those frameworks. Until now, however, the environmental footprint of AI tools has been difficult to quantify — precisely because the industry has lacked mandatory disclosure requirements.
The scale of AI's resource demand
A June 2026 study from the United Nations University (UNU), cited in the UN initiative, offers a striking baseline. According to the UNU research, data centres powering AI could consume 945 terawatt-hours of electricity annually by 2030 — nearly triple the combined annual electricity use of Pakistan, Bangladesh, and Nigeria. The same study found that AI-related water consumption could equal the basic annual domestic needs of approximately 1.3 billion people by the end of the decade, while the technology's land footprint may exceed 14,500 square kilometres.
The report also challenges the assumption that training large AI models is the primary driver of environmental impact. In fact, day-to-day usage accounts for roughly 80 to 90 per cent of total energy demand, according to the UNU findings. Generating a single AI image, for instance, can require more than a thousand times the energy of a simple text classification task.
Canadian organizations with ESG commitments or sustainability reporting obligations may soon find that their AI procurement decisions carry environmental implications they are not yet measuring. HR leaders in this position face questions about the vendors, the technologies, and the use cases they are choosing to prioritize.
No standardized environmental reporting
The UN initiative identifies a core problem that applies directly to procurement: the absence of standardized environmental reporting means organizations can’t easily compare vendors. Without disclosed data on energy sources, water consumption, and carbon output, even well-intentioned ESG teams are operating without a complete picture.
Guterres called for all AI data centres to operate on renewable energy by 2030, reported Reuters. Major technology companies have made net zero commitments, but critics have noted that surging demand from AI is making those targets harder to achieve. Some providers are pivoting toward natural gas and nuclear power to meet growing energy needs, according to Reuters — a shift with its own environmental and reputational implications for organizations whose supply chains those choices affect.
The UN's framing of this issue as a transparency problem aligns with how leading HR and people teams are approaching technology governance more broadly. Just as HR leaders have developed frameworks for evaluating AI tools on bias, data privacy, and labour market impact, environmental accountability is emerging as an additional lens.
For organizations already applying ESG criteria to procurement, adding environmental disclosure requirements to AI vendor assessments is a logical extension of existing practice. The critical gap — and one that the UN initiative is designed to close — is the absence of comparable, verified data from vendors.
A governance gap with real consequences
The UNU report notes that the environmental impacts of AI infrastructure are not evenly distributed. While the benefits are global, the costs are often concentrated in specific regions — a dynamic that raises equity questions for organizations with international operations.
More immediately for HR leaders in Canada, the report warns of a growing electronic waste challenge, with AI infrastructure projected to generate up to 2.5 million tonnes of e-waste annually by 2030. That burden is expected to fall disproportionately on lower-income countries with limited capacity for safe disposal — a supply chain ethics issue that responsible procurement frameworks are beginning to address.
“No more hidden costs,” Guterres said, according to the Associated Press. “No more shifting the burden onto those least able to bear it. It is time to come clean.”
The UNU researchers are explicit that their findings are not an argument against AI. Their call is for what they describe as a "responsible AI ecosystem," built on transparency, efficiency by design, and lifecycle accountability — principles that translate directly into the governance questions HR leaders are already grappling with.
As Canada's regulatory and corporate governance environment continues to evolve, the organizations that build environmental accountability into their AI strategies now are likely to be better positioned than those that treat it as a future consideration.
