By Ryan D’Souza
India’s rapid adoption of AI spans advanced research and development, enterprises, and startups alike, transforming everything from automation to advanced data solutions. While organisations accelerate to integrate AI into their operations, it becomes critical to address the technology’s environmental footprint, which is a growing challenge that could undermine both sustainability ambitions and long-term business resilience. Here are five critical mistakes business and technology leaders in India should avoid when scaling AI responsibly:
Overengineering AI: As organisations explore the power of advanced AI, it’s important to align model choice with the task at hand. While large-scale models unlock remarkable capabilities, which can consume 10-100x more energy per query, many applications can be addressed just as effectively with smaller, optimised models, delivering comparable results with far greater energy efficiency. Selecting the right-sized model not only optimises performance and cost but also supports sustainability goals by reducing computational overhead.
Building infrastructure
Sustainable infrastructure: AI’s carbon footprint depends as much on infrastructure as on innovation. Not all infrastructure components or data centres are created equal when it comes to energy use and carbon emissions. For example, running the same model in a coal-powered data centre versus one powered by renewables can create up to a tenfold difference in emissions. With the rapid growth of India’s AI data centre ecosystem, the integration of renewable energy and sustainable cooling technologies will be key to achieving long-term efficiency, resilience, and environmental stewardship.
Data value: Storing massive datasets indefinitely without governance policies leads to unnecessary energy consumption from data centres. Data is the fuel of AI, but poor data hygiene creates enormous hidden costs. Many organisations collect and store vast amounts of data “just in case,” without policies for data lifecycle management. Redundant, obsolete, or trivial data still requires energy to store, back up, and maintain. Implement data governance frameworks that regularly assess data value, establish retention policies, and use compression techniques. Consider whether there is a need to store raw data permanently or if processed, smaller datasets would suffice for the AI applications.
Underinvesting in people: Implementing AI without properly training your workforce or communicating how AI will augment rather than replace human roles. AI success is as much about people as it is about technology. When organisations invest in upskilling, transparency, and clear communication, they minimise resistance and ensure smooth adoption. Without this focus, employees may feel uncertain or underprepared, leading to lower utilisation, inefficiencies, and missed opportunities. By prioritising people alongside technology, businesses can make AI adoption more sustainable, both environmentally and economically.
Sustainability
Clear communication that positions AI as a collaborative tool, helping humans focus on higher-value, creative work, combined with comprehensive training programs, ensures your AI investments deliver their intended benefits while maintaining workforce engagement and productivity.
AI sustainability metrics: Treating AI sustainability as an afterthought rather than building measurement and monitoring into your AI operations from day one. Effective management begins with meaningful measurement. Most organisations have no visibility into the energy consumption or carbon emissions of their AI systems, making it impossible to identify optimisation opportunities or track progress toward sustainability goals. Implement tools and frameworks that monitor key efficiency metrics such as tokens processed per kilowatt-hour, inference requests per unit of energy consumed, model response time versus computational load, and carbon emissions per user interaction.
AI sustainability isn’t just about carbon emissions and environmental risk mitigation, but it’s smart business. The key is to embed sustainability thinking into your AI strategy from the beginning, not bolt it on afterwards. In an AI-driven world, sustainability isn’t just responsible, but it’s a strategic advantage. Organisations that adopt sustainable AI practices today will be the ones leading with efficiency, resilience, and competitive strength tomorrow.
The writer is country manager – HPC, AI & NonStop Solutions, HPE India
