Your AI Strategy Has an ESG Blind Spot

Hello H.A.I.R. Community,

Across every industry, the mandate is clear: implement AI. From HR teams automating job descriptions to marketing personalising campaigns, the rush to deploy generative AI is fundamentally reshaping the enterprise. But in this race for efficiency and innovation, a critical governance function is being left behind, creating a significant and unmeasured ESG (Environmental, Social, and Governance) blind spot: the sustainability team.

The prevailing assumption is that AI, as software, is a "virtual" resource with a negligible environmental cost. A landmark 2025 research paper from Google, "MEASURING THE ENVIRONMENTAL IMPACT OF DELIVERING AI AT GOOGLE SCALE," definitively proves this assumption is not only wrong but dangerously misleading. For any leader serious about Governance, Risk & Compliance, the findings are a call to action: your sustainability experts must have a seat at the AI strategy table from day one.

Let’s dive into it.

The Reality: From Virtual Hype to Measured Impact

For years, the environmental cost of an AI query was a matter of high-level estimates. The Google paper changes this by providing a comprehensive, full-stack measurement of its Gemini AI assistant in a live production environment. The data reveals a tangible, physical footprint for every single query.

A median Gemini text prompt was found to consume:

• 0.24 Wh of energy.

• 0.26 mL of water.

• 0.03 gCO₂e in market-based carbon emissions.

Crucially, the paper argues that the method of measurement is paramount. Its proposed "Comprehensive Approach" includes energy for host systems, essential idle machines, and full data centre overhead. This comprehensive total is 2.4 times greater than what narrower, more common measurement approaches would suggest. This is exactly the kind of methodological diligence that a sustainability professional is trained to scrutinise and a detail that an IT or business unit, focused solely on performance, might overlook.

The Scale Problem: From Single Prompts to an Enterprise Footprint

Critics may dismiss these figures as trivial—equivalent to just "five drops of water" or less energy than nine seconds of television. This misses the point entirely, a classic fallacy of scale.

To put this into perspective, let's consider the average Fortune Global 500 company, which employs approximately 140,200 people. Assuming a moderate usage of 109 prompts per user per month (calculated using OpenAI’s latest statistics - 700mn users per week and 2.5bn prompts per day), the cumulative environmental footprint becomes a significant figure on any corporate sustainability report.

Here is the projected monthly impact at various levels of adoption for such a company:

These are not trivial numbers. They represent a new and material source of energy consumption, carbon emissions, and water usage that must be measured, managed, and reported. A sustainability team is uniquely equipped to model this scaled impact and integrate it into the company's overall environmental accounting.

The Due Diligence Imperative: Not All AI Is Created Equal

Perhaps the most critical reason for involving sustainability teams is that the environmental efficiency of AI models is not standard across the industry. Google's low impact figures are not an accident; they are a hard-won engineering achievement.

The research highlights a staggering 44x reduction in total emissions per prompt between May 2024 and May 2025. This was the result of relentless, continuous optimisation across the entire stack, from creating more efficient model architectures to co-designing custom hardware and procuring clean energy.

This proves that efficiency is a key differentiator among AI providers. A company procuring an AI solution without this lens risks partnering with a vendor whose models are orders of magnitude less efficient, creating an unnecessary and significant environmental liability.

Your sustainability team must be embedded in the AI procurement process to ask the critical GRC questions:

• What is your energy consumption and carbon footprint per query?

• What is your measurement methodology, and is it comprehensive, including idle capacity and data centre overhead?

• Can you demonstrate year-on-year efficiency gains and a clear roadmap for further reductions?

An AI vendor who cannot answer these questions is a sustainability risk.

Turning a Blind Spot into a Strategic Advantage

The implementation of AI is a defining moment for modern business. Treating it as a purely technological rollout is a mistake that will leave your organisation exposed to regulatory, reputational, and financial risks.

By embedding sustainability leaders into the process from the very beginning, you transform a potential ESG blind spot into a strategic advantage. You ensure that your company is not only harnessing the power of AI but doing so responsibly, efficiently, and transparently. In an era where stakeholders demand both technological innovation and environmental accountability, you cannot afford to achieve one at the expense of the other.

Here's how H.A.I.R. can help you put the AI in HR:

1. H.A.I.R. Newsletter: get authoritative, pragmatic, and highly valuable insights on AI in HR directly to your inbox. Subscribe now.

2. EU AI Act QuickScore Assessment: understand your organisation's EU AI Act readiness in minutes and identify key areas for improvement. Take your QuickScore here.

3. Advisory Services: implement robust AI Governance, Risk, and Compliance (GRC) with our 12-month programme designed for HR and Talent Acquisition leaders. Contact us for a consultation.

4. Measure Your Team's AI Readiness with genAssess: stop guessing and start measuring your team's practical AI application skills. Discover genAssess.

Until next time,

Martyn

H.A.I.R. (AI in HR) 

Putting the AI in HR. Safely.