Exacerbating Climate Change
the growing field of generative AI, which brings with it direct and severe impacts on our climate: generative AI comes with a high carbon footprint and similarly high resource price tag, which largely flies under the radar of public AI discourse. Training and running generative AI tools requires companies to use extreme amounts of energy and physical resources. Training one natural language processing model with normal tuning and experiments emits, on average, the same amount of carbon that seven people do over an entire year.121'
ENTITY
2 - AI
INTENT
2 - Unintentional
TIMING
3 - Other
Risk ID
mit526
Domain lineage
6. Socioeconomic and Environmental
6.6 > Environmental harm
Mitigation strategy
1. Establish Standardized Metrics and Mandatory Disclosure for Environmental Footprint Mandate the development and adoption of standardized, auditable procedures for measuring and disclosing the environmental impact—including operational and embodied carbon emissions, Water Usage Effectiveness (WUE), and electronic waste—across the entire lifecycle of AI models (training, fine-tuning, and inference). 2. Implement Algorithmic and Model Optimization for Computational Efficiency Incentivize and regulate the development and preferential deployment of resource-efficient AI paradigms, such as smaller, domain-specific Small Language Models (SLMs) and specialized accelerators (e.g., TPUs, FPGAs), in place of massive, general-purpose models to reduce the per-query energy and resource demand during inference. 3. Mandate a Transition to Sustainable Data Center Infrastructure Enforce regulatory requirements for data centers hosting AI workloads to prioritize verifiable 100% carbon-free or renewable energy procurement, implement advanced water-recycling and closed-loop cooling systems, and institute robust circular economy strategies for hardware reuse and e-waste mitigation.