India must focus on AI and its environmental impact, Pg8

Key Highlights:

• The global ICT industry contributes 1.8%-3.9% of global greenhouse gas (GHG) emissions.
• A UNEP note in September 2024 warned that AI servers could use 4.2 to 6.6 billion cubic meters of water by 2027, exacerbating water scarcity.
• Training a single Large Language Model (LLM) can generate approximately 3,00,000 kilograms of carbon emissions.
• UNESCO released its "Recommendation on the Ethics of Artificial Intelligence" in 2021, emphasizing the negative environmental impacts of AI.
• The US and EU have proposed legislation like the Artificial Intelligence Environmental Impacts Act of 2024 and resolutions on harmonized AI rules.

Detailed Insights:

• The development of AI algorithms has environmental costs, including an increased carbon footprint, which intensifies climate change.
• A Google report in August 2025 claimed low electricity consumption for AI prompts, but it faced criticism for incomplete conclusions.
• Training a single large AI model can emit over 6,26,000 pounds of carbon dioxide, equivalent to the emissions of five cars over their lifespan.
• Using ChatGPT for a request consumes 10 times more energy than a Google search, highlighting the environmental impact of AI virtual assistants.
• India needs to measure the environmental impacts of developing and deploying AI models, extending the scope of Environmental Impact Assessments (EIA).
• The government should establish measuring standards for AI's impact, involving tech companies, think tanks, and NGOs to build consensus.
• Collecting data using sustainability metrics like GHG emissions, energy, water, and natural resource consumption is essential for assessing AI's environmental costs.
• Integrating the environmental impact of AI into ESG disclosure standards, inspired by the EU's Corporate Sustainability Reporting Directive (CSRD), is vital.
• Sustainable practices like deploying pre-trained models, using renewable energy for data centers, and reporting AI-specific estimates can mitigate adverse environmental impacts.

Key Concepts Involved:

• Artificial Intelligence (AI): Simulation of human intelligence processes by computer systems.
• Greenhouse Gas (GHG): Gases that trap heat in the atmosphere and contribute to global warming.
• Environmental Impact Assessment (EIA): Evaluation of the environmental consequences of a plan, policy, program, or project.
• ESG Disclosure Standards: Reporting of environmental, social, and governance factors by companies.