Companies face pressure to cut carbon footprints while maintaining performance, and a new approach focuses on optimizing employee interactions with AI systems, as prompt wording directly affects energy consumption. Scott Lane, Founder and CEO of Speeki, a sustainability firm with an AI-powered audit platform, said his firm has more AI agents than humans, forcing an evaluation of environmental impact. He noted control over two choices: hosting providers and prompt writing. Cloud services are commoditised, enabling selection based on environmental performance, while prompt engineering offers an immediate user-controlled lever for cutting emissions. Lane said cutting out pleasantries like "please" and "thank you" is a simple way to lower AI energy consumption. According to a 2025 technical paper on Google's Gemini models, a median text prompt consumes 0.24 watt-hours of energy, emits 0.03g of carbon dioxide, and uses five drops of water, equating to less than nine seconds of watching television. A study titled Green Prompt Engineering for Sustainable Generative AI by researchers at Boston University, the University of California, and the University of Nottingham found optimized prompting reduces LLM inference energy and carbon dioxide emissions by up to 48%, cutting power across systems including OpenAI's GPT-4o, Cohere, and Mistral AI's Mistral 7B. Lane expressed optimism that technological innovation will keep pace with infrastructure demands, betting on rapid equipment performance advancements, with AI likely helping produce more efficient and sustainable technologies.