The potential of AI – can it help or is it all hype when it comes to sustainability?

AI has become a polarising topic with a lot of myth surrounding it. For some, it’s a breakthrough technology that could help solve our most complex sustainability challenges. For others, it’s overhyped, energy-hungry, and riddled with ethical concerns. As with most things, the reality sits somewhere in between.

But alongside the optimism, there is growing scepticism. Could AI become the next big sustainability challenge? Training and running advanced AI systems demands huge amounts of energy, with companies like Microsoft and Google, both long-time sustainability leaders, reporting significant increases in carbon emissions linked to AI investment. Cooling enormous data centres requires water-intensive processes, while the rapid obsolescence of hardware adds to mounting e-waste and undermines circular economy principles.

Beyond the environmental footprint, questions of accuracy and bias raise further concerns. Distorted sustainability information can mislead and algorithms that reflect or amplify inequalities risk diverting resources away from underserved communities or underrepresented stakeholders. At the same time, the spread of automation fuels anxiety about job losses, even though history suggests that new technologies tend to reshape roles rather than simply replace them.

There are further reasons to be cautious, in its current state, 95% of generative AI implementations “have no measurable impact on P&L”, according to MIT, with flawed integration often to blame. That gap between potential and performance matters, especially for sustainability, where execution is everything. Increasingly, ESG is being reframed not as a compliance exercise, but as a lever for value creation and protection, helping companies reduce risk, build resilience and unlock growth. When deployed well, AI could accelerate this shift – driving sharper insights, smarter resource allocation and greater accountability. But when it falls short, it risks eroding trust, inflating costs and deepening existing sustainability challenges.

The real question is, for growing businesses, can AI help resource-constrained teams work more efficiently when it comes to sustainability, or will it simply become another drain on already limited time, budget and resources? To explore this, we look through three lenses: what AI can do today, what’s emerging on the near horizon, and what a more speculative future might hold.

Today’s AI tools for sustainability have significant limitations

Most of today’s tools fall under Narrow AI – algorithms designed for specific tasks. Large Language Models, like ChatGPT, are mainly focused on content generation. For ESG, they can process vast amounts of sustainability data quickly, generate reports, or translate technical content into accessible formats for stakeholders.

But there are significant limitations. LLMs still rely on human-prepared data, and issues like “hallucinations” mean outputs can sound convincing while being factually wrong. For ESG reporting, where compliance failures carry big consequences, these tools remain unsuitable without human oversight. The time saved on report writing can easily be lost in fact-checking, meaning AI is still more of a support act, than a game-changer here.

Emerging AI is beginning to shake up sustainability

The next wave is Agentic AI – systems that go beyond generating content to making decisions and optimising operations. These tools don’t rely singularly on human prompts and are beginning to show real promise across specific or bespoke sustainability challenges.

Many companies lack the data and resources to deliver on their otherwise credible Net Zero strategies and science-based targets. Salesforce’s AI to track emissions, model decarbonisation pathways and provide predictive insights. It can help organisations identify high-impact reduction opportunities and shift climate strategy from static reporting to dynamic decision-making.

Investors, insurers and consumers are demanding climate resilience, but assessing risk from floods, heatwaves, or fires is costly and highly technical. provides AI-driven analysis, enabling lenders and investors to integrate climate risk into financing decisions. This kind of pragmatic sustainability data can help businesses reduce risk, attract investment and grow.

As businesses rush to showcase their sustainability credentials, many overlook the hidden footprint of their digital operations. , uses AI to monitor web emissions, which is the carbon generated when websites consume energy through servers and data transfers, and provide reduction strategies. It helps smaller businesses cut hidden emissions and avoid greenwashing, showing that AI is not just for big tech; smaller players can use it to stay credible too.

Each of these examples point to the same trend, as AI starts to become increasingly designed for specific ESG practices, it offers efficiencies that can accelerate progress, helping guide smarter decisions with less time, budget and resources. However, the immediate value of these emerging tools depends on context; where there is a clear use case, AI can be a powerful enabler, but without one, it risks becoming a distraction or failing to deliver on its promise altogether.

The AI of tomorrow could be revolutionary

Looking further ahead, speculation often turns to Artificial General Intelligence (AGI) – a form of AI that, in theory, could replicate a wider set of cognitive abilities such as learning, reasoning, and problem-solving – with multiple AI tools collaborating in real time. It’s worth stressing that AGI is not yet here, and it may never arrive. But if it does, the implications for sustainability could be significant.

Rather than replacing human judgement or jobs, the real potential lies in how AGI could act as an accelerator and partner that models complex climate scenarios, tests Net Zero pathways, or helps leaders weigh trade-offs across social, environmental, and economic priorities. It could reimagine how resources are managed at scale or, if democratised, provide faster access to sustainability knowledge, making it easier for people and businesses to make informed decisions. But if concentrated in the hands of a few with wealthier companies, countries or people able to adapt and respond while smaller players are left behind.

How should we consider AI for ESG?

AI today is already offering useful tools that can streamline aspects of sustainability. Tomorrow’s wave promises even more, while the more distant future raises both worrying risks and exciting possibilities. The truth is that AI is neither saviour nor villain. It is a powerful, imperfect tool, one that will be defined by how we choose to use it.

For businesses wanting to scale up their sustainability efforts, the challenge is cutting through the AI hype and focusing on where it can add real value. That requires experimenting and implementing AI responsibly through making systems smarter, smaller and stronger to improve energy efficiency, limit waste technology, get ahead on standards and consider the ethics of automation by prioritising reskilling and role evolution over replacement of employees. For smaller businesses, starting with lighter, well-defined AI tools can help cut costs, save time and lower emissions, while embedding ethical guardrails from the outset.