Artificial Intelligence's hidden cost: Why responsible use matters

Artificial intelligence is rapidly transforming publishing, media, art and almost every other sphere of human activity. Yet, as organisations rush to embrace its possibilities, an equally important question is beginning to emerge: what is the environmental price of this technological revolution?

The enormous computing infrastructure behind AI requires vast quantities of electricity and water, while constant hardware upgrades contribute to the growing problem of electronic waste.

At the ‘Beyond Tomorrow’ conference organised by Kerala Startup Mission (KSUM) at the Kalamassery Digital Hub, this less-discussed dimension of AI came into sharp focus.

Speaking on ‘Publishing and Art in the Age of AI’, Mathrubhumi Director-Operations Devika MS warned against the indiscriminate use of artificial intelligence and called for a more responsible approach to the technology.

The debate around AI, she noted, has largely concentrated on familiar concerns such as misinformation, copyright violations and the future of jobs. The environmental consequences of AI, however, have received far less attention.

"Even a small data centre uses enough electricity to power about 10,000 homes," Devika said, underlining the scale of energy consumption associated with the infrastructure supporting AI. The concern becomes even more significant when the demand created by increasingly sophisticated AI systems is considered.

Training and operating large AI models require powerful servers that must run continuously, creating an energy burden that extends far beyond the computers or smartphones through which ordinary users interact with these systems.

The environmental footprint does not end with electricity. Data centres also require enormous quantities of water, particularly for cooling the equipment that generates heat during intensive computing operations. Devika pointed to an estimated consumption of around five million gallons of water a day for a data centre, describing the figure as far from insignificant.

At a time when many parts of the world are already confronting water stress and the effects of climate change, the rapid expansion of computing infrastructure raises questions about how technological progress should be balanced with environmental responsibility.

Also Read: Kerala’s first AI data centre to come up in Kozhikode; ₹1,200-crore IT project planned

The issue is becoming increasingly relevant as AI moves from being a specialised technology to an everyday tool. Generative AI can produce text, images, audio and video within seconds, making it tempting to employ the technology for even the smallest tasks. But the apparent simplicity of clicking a button masks the physical infrastructure required to process every request. Each AI interaction ultimately depends on servers, networks, cooling systems and electricity.

Devika, therefore, argued for restraint rather than rejection. The question, she suggested, is not whether society should stop using AI, but whether it should use the technology thoughtfully and only when there is a genuine need.

"Instead of using AI for even the smallest things we can do, we should try to use AI responsibly only where necessary," she said.

That philosophy is also reflected in Mathrubhumi’s own approach to the technology. Devika explained that the organisation is already using AI in areas such as proofreading, personalisation and content insights. These applications demonstrate how AI can support media organisations without necessarily replacing the human judgement that remains central to journalism and publishing.

Mathrubhumi is also exploring the use of AI to make better use of its extensive archival content. For a media organisation with decades of journalism and cultural documentation, AI could help organise, search and connect archival material in ways that were previously difficult to achieve. Old photographs, articles, reports and other records could potentially become more accessible to journalists, researchers and readers through intelligent search and classification systems.

However, such applications need to be accompanied by awareness of their environmental cost. “We should also not overuse AI as data centres will be consuming a lot of water,” Devika observed, linking excessive dependence on computing infrastructure to wider climate and environmental concerns.

Also Read: Did Anthropic's AI create fake identities to fool people? UK test raises fresh concerns

The discussion at Beyond Tomorrow placed these concerns within a broader conversation about creativity. Sajoy Roy, Co-Chair of the FICCI Tourism and Culture Committee, said during the inaugural session that AI could not replace the fundamental idea of creativity. For him, creativity remains essential not only to artistic expression but also to generating employment opportunities for young people and addressing the challenges created by technological change.

The point is particularly significant at a time when AI is increasingly capable of producing content that appears creative. Text can be generated within seconds, images can be created from descriptions, music can be synthesised and video production is becoming increasingly automated. Yet the ability to generate an output is not necessarily the same as possessing imagination, cultural understanding or lived experience.

Human creativity is shaped by memory, emotion, observation, social relationships and an understanding of context. These qualities remain central to journalism and the arts. AI can assist creators, accelerate certain processes and provide new possibilities, but the responsibility for deciding what deserves to be created and why ultimately remains human.

The environmental question adds another layer to this debate. If AI is treated as an automatic solution for every task, the resulting demand for computing power could grow without meaningful consideration of its ecological consequences. Responsible AI, therefore, must involve not only questions of accuracy, copyright, transparency and ethics but also questions of energy, water and waste.

The hardware required to operate AI systems presents another challenge. As servers and processors are upgraded to meet the demands of increasingly powerful technologies, older equipment eventually becomes electronic waste. The rapid pace of technological development could therefore create a cycle in which computing infrastructure is continually replaced, adding to the global burden of e-waste.

This makes the idea of ‘progress’ more complicated. A technology may be innovative and efficient in one sense while creating new environmental pressures elsewhere. The real measure of technological advancement may consequently depend on whether its benefits can be achieved without transferring unacceptable costs to the environment and future generations.

The Beyond Tomorrow conference brought together voices from the worlds of media, art, technology, public administration and entertainment to examine precisely these changes. Malayala Manorama Executive Editor and Director Jayant Mathew, former Kerala Lalithakala Akademi Chairperson Murali Cheeroth, AI expert Sleeba Paul and KSUM Senior Fellow Dr Tanya Abraham joined Devika in the discussion on publishing and art.

State Civil Supplies Corporation Managing Director VM Jayakrishnan delivered the keynote address, while KSUM COO Tom Thomas, EEMA Secretary Harish Babu, CRISIL Intelligence Urban Infrastructure and Public Finance Practice Leader Suraj Iyengar, film critic and documentary filmmaker Dr CS Venkiteswaran and film star Raveendran were among those who participated in discussions on various subjects.

The range of speakers reflected the breadth of the AI debate itself. Artificial intelligence is no longer merely a technology-sector issue. It has become a question for publishers, artists, filmmakers, governments, businesses, educators and society at large.

For the creative industries, the challenge is particularly delicate. AI offers powerful tools that can improve productivity and open new avenues for experimentation. At the same time, excessive dependence can dilute human involvement, raise questions about originality and consume resources that remain largely invisible to the end user.

The future, therefore, may not belong to those who use the most AI, but to those who understand when and how to use it wisely. For media organisations such as Mathrubhumi, that means exploring AI where it adds genuine value while retaining human judgement at the heart of journalism. For society, it means recognising that every seemingly effortless digital service has a physical footprint.

The conversation around AI has often been framed as a contest between humans and machines. The more urgent contest may be between technological ambition and environmental responsibility. AI can help humanity solve complex problems, but it must not become a new source of problems that are ignored because they remain hidden behind screens and algorithms.

As the AI era advances, responsible use will require more than ethical guidelines. It will require awareness of the electricity consumed, the water used, the hardware discarded and the environmental consequences of every new layer of digital infrastructure. The question is no longer simply what AI can do. It is also whether everything AI can do needs to be done.

That distinction could ultimately determine whether the promise of artificial intelligence becomes a sustainable part of tomorrow—or an environmental burden that future generations will have to confront.