AI is transforming the way we work in market research, but this progress comes at a cost. Every AI prompt, dataset, and model requires energy and, importantly, water. Behind the scenes, vast data centres run around the clock to power these systems and keep them cool.

It’s estimated ChatGPT used 14.46bn kWh in a single year – more than the annual consumption of 117 countries[ 1 ].  Cooling infrastructure for this level of computing therefore demands significant water use, with researchers at the University of California estimating a typical ChatGPT session consumes half a litre of water[ 2 ].

The hidden cost of intelligent tools

AI feels effortless. A few prompts, and the solutions appear. Yet beneath that lies an immense physical infrastructure: networks of servers drawing electricity, generating heat and consuming water at scale to ensure they stay cool.

These costs rarely appear in our day-to-day thinking, but their impact ripples far beyond the digital space. Data centres rely on local water supplies, adding pressure to regions already struggling with scarcity. As AI tools become more embedded in our research, it’s easy to forget that each prompt quietly consumes resources.

As researchers, we are trained to ask questions, to challenge assumptions and look beneath the surface. That same curiosity can help us approach innovation differently. We need to see AI as part of a larger ecosystem that deserves our attention and care.

Understanding AI’s footprint does not mean rejecting progress. It means being bold and intelligent about how we move forward. We can ask not only what AI can do, but what it should do, and how we can use it responsibly. True innovation recognises both opportunity and cost; and acts with awareness of both.

Balancing innovation with responsibility

Awareness, transparency and education are essential to establishing balance when using AI. These help researchers, clients and organisations understand not only what AI can do, but what it costs to use.

As an industry built on curiosity and understanding cause and effect, we can’t ignore these hidden consequences. When we first learned about the environmental impact of AI, we were genuinely shocked – especially given how often we, and the wider industry, rely on these tools, with usage only continuing to grow.

It sparked conversations across our organisation, prompting us to reflect on our own AI footprint, explore our options, and think carefully about what meaningful action could look like for us. The scale of water use, in particular, felt impossible to overlook, especially at a time when water scarcity is a global challenge – that’s why we chose to partner with WaterAid, committing £20 for every completed project to support global water security. It’s a small but intentional step towards giving back to the resource that so much of our innovation quietly depends on.

We are also opening these conversations within our teams, including training researchers to use AI thoughtfully by choosing efficient prompts, for example, and being aware of the environmental implications of digital actions.

By supporting projects that bring clean water and resilience to communities worldwide, we are recognising the link between technological progress and environmental stewardship. The more we understand our impact, the better we can make choices that matter, both professionally and personally.

We know this alone is not the solution. True change requires collective effort across industries and systems. But small, meaningful actions matter. They set a standard, spark dialogue and help embed awareness where it counts, in the everyday decisions researchers make.

Redefining what intelligent insight really means

AI is transforming how we find and understand insight, but true intelligence in research has never just been about technology. It is about how thoughtfully we use the tools we have in a human-led way.

As innovation accelerates, we need to keep asking the right questions. Are we using AI to make work smarter, or simply faster? Are we improving understanding, or just increasing output? Intelligent insight means knowing the difference.

It is about awareness as much as accuracy; recognising intelligent insight combines curiosity with care and responsibility. The best research doesn’t just explain what people do; it considers the impact of how we get there.

Researchers are already equipped to lead this. We know how to question, test and interpret complexity, and those same skills can guide how we use AI responsibly. When we balance creativity with conscience, we create insights that serve both progress, people and planet.

Emily Jackson is research manager at 2CV

Reference:

[ 1 ] Source: https://www.businessenergyuk.com/knowledge-hub/chatgpt-energy-consumption-visualized/

[ 2 ] Source: https://arxiv.org/abs/2304.03271