A United Nations entity has warned of the rapidly increasing environmental burdens imposed by the accelerated expansion of artificial intelligence technologies. It called upon governments, companies, and investors to integrate environmental considerations into all stages of this technology's development, aiming
to ensure its sustainable growth without exacerbating pressure on vital natural resources. This warning emerged from a report issued by a UN University Institute for Water, Environment and Health (UNU-INWEH) in Canada. The report addressed the escalating environmental impacts of artificial
intelligence, ranging from the production of electronic chips and the extraction of essential minerals, through the immense consumption of energy and water, to the challenges of electronic waste. The report clarified that data centers, which form the backbone of artificial intelligence
systems, deplete vast quantities of electricity and water required for cooling operations, in addition to needing extensive land areas. Furthermore, the cumulative impact of billions of daily user interactions, from messages and queries to generated images, significantly contributes to the
exacerbation of this environmental burden. The report indicated that the global artificial intelligence market is projected to grow 25-fold, reaching approximately five trillion dollars by 2033. It cautioned that this rapid expansion of technology could lead to the accumulation of social,
economic, geopolitical, and environmental impacts that will be difficult to address once the associated systems and investments become entrenched. Although public discussions often focus on the risks associated with artificial intelligence in aspects such as bias, privacy, misinformation, and the job
market, the report considered the technology's environmental impact to be an aspect that has not received sufficient attention, particularly concerning environmental justice and the distribution of burdens among different nations. Regarding energy consumption, scientists revealed that artificial intelligence-related workloads will account
for about 20% of the total electricity consumption in data centers by 2025, with this percentage expected to rise to 40% by the end of the current decade. According to the report, electricity consumption associated with artificial intelligence could, by 2030,
reach a level sufficient to meet the residential electricity needs of sub-Saharan Africa's 1.3 billion inhabitants for over two years. The total demand for electricity from data centers is also expected to nearly double by the same year. The report added
that generating these quantities of electricity could result in emissions of approximately 400 million tons of carbon dioxide, which would require planting nearly 6.7 billion trees over ten years to offset. As for water consumption, it is expected to approach 9.3
trillion liters by 2030, a quantity sufficient to meet the drinking water needs of the world's 8.1 billion people for approximately 1.6 years. The report also projected that artificial intelligence infrastructure would produce up to 2.5 million tons of electronic waste
annually by 2030. UN experts questioned the prevailing belief that operating data centers with renewable energy necessarily makes them environmentally friendly. They explained that a reduction in carbon emissions does not negate other impacts associated with the consumption of water, land,
and natural resources. The report also observed a widening digital divide between countries that develop and manage artificial intelligence systems and those countries that merely consume these technologies. It noted that only 32 countries host specialized cloud infrastructure for artificial intelligence,
while 90% of these capabilities are concentrated in the United States and China. It added that over 150 countries lack sovereign artificial intelligence infrastructure, while some excluded nations receive fewer benefits in exchange for significant environmental burdens linked to mineral extraction
and handling electronic waste. To ensure sustainable growth for artificial intelligence, the report urged governments to align technology infrastructure plans with energy policies, water management, and land use. It also encouraged companies and developers to consider environmental impact when designing models
and improving their efficiency. It called on users to choose less energy-intensive models when they meet their needs. Furthermore, it requested data center operators and investors to treat electricity, water, land consumption, and carbon emissions as fundamental factors in future
decision-making.
