The chapter discusses the environmental costs of generative AI, highlighting the energy and water consumption associated with its operation. OpenAI CEO Sam Altman recently acknowledged the industry's impending energy crisis, emphasizing the need for breakthroughs like nuclear fusion to address the issue. However, experts argue that nuclear fusion will not significantly contribute to decarbonization efforts. The chapter also notes that generative AI systems require substantial amounts of fresh water for cooling and electricity generation, with data centers using a significant portion of local water resources. Despite the urgency, the industry has been slow to adopt sustainable practices, and accurate data on environmental impacts is often lacking. To address these issues, the chapter suggests a multifaceted approach involving industry, researchers, and legislators. This includes prioritizing energy efficiency, developing sustainable hardware and algorithms, using renewable energy, and implementing mandatory environmental reporting and impact assessments. The introduction of the Artificial Intelligence Environmental Impacts Act of 2024 by US Democrats is seen as a positive step, but more comprehensive measures are needed to truly mitigate the environmental impacts of AI.The chapter discusses the environmental costs of generative AI, highlighting the energy and water consumption associated with its operation. OpenAI CEO Sam Altman recently acknowledged the industry's impending energy crisis, emphasizing the need for breakthroughs like nuclear fusion to address the issue. However, experts argue that nuclear fusion will not significantly contribute to decarbonization efforts. The chapter also notes that generative AI systems require substantial amounts of fresh water for cooling and electricity generation, with data centers using a significant portion of local water resources. Despite the urgency, the industry has been slow to adopt sustainable practices, and accurate data on environmental impacts is often lacking. To address these issues, the chapter suggests a multifaceted approach involving industry, researchers, and legislators. This includes prioritizing energy efficiency, developing sustainable hardware and algorithms, using renewable energy, and implementing mandatory environmental reporting and impact assessments. The introduction of the Artificial Intelligence Environmental Impacts Act of 2024 by US Democrats is seen as a positive step, but more comprehensive measures are needed to truly mitigate the environmental impacts of AI.