Sustainable artificial intelligence: why ESG risks are a growing concern for businesses
The relationship between artificial intelligence and sustainability has become one of the most pressing topics in European boardrooms. As organisations rush to adopt AI-driven tools, a critical question emerges: is your company’s use of intelligenza artificiale sostenibile — sustainable AI — actually sustainable? A longitudinal analysis spanning 25 years of media coverage reveals that public sentiment toward AI varies dramatically across environmental, social, and governance dimensions. For European SMBs, ignoring these shifts could mean walking into a reputational minefield.
The paradox is striking. AI promises to optimise energy grids, reduce waste, and accelerate the green transition. Yet the technology itself consumes enormous quantities of electricity and water, generates significant carbon emissions, and raises uncomfortable questions about labour displacement. Understanding both sides of this equation is no longer optional — it is a strategic imperative for any business operating under the European regulatory umbrella.
The environmental paradox of AI: efficiency gains versus resource consumption
One of the most revealing findings from recent sentiment research is the stark contradiction at the heart of AI’s environmental narrative. On one hand, AI systems help companies monitor emissions, optimise logistics, and improve energy efficiency. On the other, training a single large language model can emit as much carbon dioxide as five cars over their entire lifetimes — a figure first highlighted by researchers at the University of Massachusetts Amherst and repeatedly confirmed by subsequent studies.
The numbers have only grown more alarming. According to the International Energy Agency (IEA), data centres worldwide consumed approximately 460 TWh of electricity in 2022, and this figure is projected to exceed 1,000 TWh by 2026 — roughly equivalent to Japan’s total electricity consumption. A significant share of that growth is driven by AI workloads. Google’s own environmental reports acknowledged a 48% increase in greenhouse gas emissions between 2019 and 2023, largely attributed to expanding AI infrastructure.
For SMBs, these macro-level statistics translate into a concrete risk. When your company adopts AI tools hosted on hyperscale cloud platforms, you inherit a share of that environmental footprint. European customers and partners are increasingly aware of this. A 2024 Eurobarometer survey found that 77% of EU citizens consider climate change a very serious problem, and consumer purchasing decisions are shifting accordingly.
The practical implication is straightforward: businesses that cannot articulate how their AI usage aligns with sustainability goals will face harder questions from clients, regulators, and investors. This is especially true in Italy, where the National Recovery and Resilience Plan (PNRR) ties significant funding to green transition milestones.
AI e sostenibilità aziendale: how public sentiment shapes reputational risk
The 25-year media analysis referenced in the source study reveals a crucial insight — public perception of AI is not monolithic. Sentiment differs significantly depending on whether the discussion focuses on environmental impact, social consequences, or governance frameworks.
Environmental sentiment toward AI has grown increasingly negative over the past five years, driven by high-profile reports on energy consumption and water usage. Social sentiment is more mixed: people appreciate AI’s potential to improve healthcare and accessibility, but worry about job losses and algorithmic bias. Governance sentiment, meanwhile, tends to be cautiously positive, especially in Europe where the AI Act and related regulations are seen as necessary guardrails.
This fragmentation creates a complex landscape for rischi reputazionali intelligenza artificiale. A company might earn praise for deploying AI to improve customer service while simultaneously drawing criticism for the carbon footprint of its AI infrastructure. The reputational risk is not just about whether you use AI, but how you communicate about it and what measures you take to mitigate its downsides.
Italian and European SMBs are particularly exposed. Unlike large multinationals with dedicated sustainability teams and sophisticated PR departments, smaller businesses often lack the resources to manage these narratives proactively. Yet they face the same scrutiny — especially those operating in B2B supply chains where larger clients now require ESG compliance from their vendors.
Companies that invest in robust cybersecurity and compliance frameworks are better positioned to address these challenges, because the organisational discipline required for cybersecurity governance overlaps significantly with ESG reporting and AI risk management.
ESG intelligenza artificiale: the regulatory pressure is real
The European Union has established itself as the global leader in AI regulation. The EU AI Act, which entered into force in August 2024 with a phased implementation timeline, introduces risk-based classifications for AI systems and imposes specific obligations on providers and deployers. But the AI Act does not operate in isolation.
It sits alongside the Corporate Sustainability Reporting Directive (CSRD), which requires companies above certain thresholds to disclose detailed environmental and social impact data — including the impact of their technology choices. The European Sustainability Reporting Standards (ESRS) explicitly mention digital technologies as a factor in environmental reporting.
For SMBs in the EU, the regulatory convergence between AI governance and sustainability reporting means that ESG intelligenza artificiale is not an abstract concept. It is becoming an auditable requirement. Companies that supply goods or services to larger enterprises subject to CSRD will be asked to provide sustainability data about their operations, including their use of AI and cloud services.
This dovetails with other compliance frameworks already familiar to European businesses. The NIS2 Directive, for example, requires organisations in essential and important sectors to implement comprehensive risk management measures. Many of the governance processes mandated by NIS2 — risk assessment, incident reporting, supply chain security — mirror what is needed for responsible AI deployment and ESG reporting.
The message for business owners is clear: compliance is converging. Building isolated silos for cybersecurity, AI governance, and sustainability reporting is inefficient and risky. An integrated approach that addresses all three domains through a unified governance framework will save time, reduce costs, and provide better protection against regulatory penalties.
Practical steps for European SMBs to manage AI sustainability risks
Understanding the impatto ambientale AI is important, but what can a small or medium-sized business actually do about it? Here are concrete actions that can reduce both environmental footprint and reputational exposure.
Audit your AI and cloud usage
Start with a clear inventory of where AI is used in your operations, which cloud providers host those services, and what their sustainability commitments are. Major cloud providers publish sustainability reports — review them. Prefer providers that operate on renewable energy and offer carbon-neutral or carbon-negative commitments. If your IT infrastructure relies on legacy on-premises systems, consider whether a well-planned cloud migration could actually reduce your overall energy consumption.
Choose AI models proportionate to your needs
Not every task requires a massive large language model. Smaller, task-specific models consume a fraction of the energy and can often deliver equivalent or superior results for focused applications. When evaluating AI tools, ask vendors about model size, inference costs, and whether they offer optimised versions for common tasks. Right-sizing your AI usage is both an environmental and a financial win.
Integrate AI governance into your existing compliance framework
If you have already invested in cybersecurity governance — risk assessments, policy documentation, incident response plans — extend those frameworks to cover AI. Document how AI systems are selected, deployed, monitored, and decommissioned. Record the environmental considerations in your procurement decisions. This creates the paper trail you will need when CSRD-related requests arrive from your supply chain partners.
Communicate transparently
Public sentiment research shows that companies perceived as honest about trade-offs earn more trust than those that make vague sustainability claims. If you use AI, say so. If your AI usage has an environmental cost, acknowledge it and explain what you are doing to mitigate it. Greenwashing — or its emerging cousin, “AI-washing” — is one of the fastest ways to destroy trust with European consumers and regulators.
Monitor the regulatory landscape
The EU’s approach to AI and sustainability regulation is still evolving. The AI Act’s full provisions will not be fully applicable until 2027. New guidance on the intersection of AI and ESG reporting is expected from the European Financial Reporting Advisory Group (EFRAG) and national supervisory authorities. Stay informed, or work with advisors who can help you anticipate changes rather than react to them.
The bottom line: sustainability is a strategic AI decision
The era when artificial intelligence could be adopted without considering its environmental and social implications is over — at least in Europe. The convergence of the AI Act, CSRD, NIS2, and evolving public sentiment means that AI e sostenibilità aziendale is now a board-level concern, not a marketing afterthought.
For European SMBs, this is both a challenge and an opportunity. Companies that proactively address the sustainability dimensions of their AI usage will differentiate themselves in the market, strengthen their supply chain relationships, and reduce their exposure to regulatory risk. Those that ignore these trends risk finding themselves on the wrong side of both public opinion and the law.
The key is to start now, start small, and build systematically. Audit your current AI footprint, integrate AI governance into your broader compliance framework, and communicate honestly with your stakeholders. Sustainable AI is not about avoiding the technology — it is about using it responsibly and being able to prove it.