For years, we’ve seen environment, sustainability, and governance (ESG) shape discourse in corporate boardrooms, as organizations seek to address long-term challenges tied to their activities. While it hasn’t been a panacea – global environmental footprints continue to rise every year – the importance of ESG is clear. In addition to driving positive societal change, it can strengthen companies’ reputations, attracting investors and customers who value sustainability and ethical practices.
Despite its considerable energy requirements, AI has so far evaded the crosshairs of ESG reporting frameworks, for several reasons. For one, the technology’s growth has simply been too fast for ESG reporting practices to keep up. Moreover, because the technology often relies on third-party cloud services, it has been difficult to ensure that the growing resource requirements of generative AI – including electricity, water, data center infrastructure, and all the rare elements used to build them – are properly reflected in sustainability reports.
As GenAI adoption increases, there is no doubt that its environmental footprint, too, will increase. Less certain, at least for now, is just how regulators will respond in particular as organizations back away from commitments that slow growth in other key areas, such as diversity, equity and inclusion (DEI). Will they temporarily turn a blind eye to encourage AI adoption so that organizations – and society at large – can reap its full benefits faster? Or will they instead try to catch up with the market and standardize how the impacts of AI are captured in corporate ESG reports?
Amid this uncertainty, the argument in favor of the cautionary principle, developing enterprise GenAI applications with ESG principles in mind, is strong. But what do ESG-focused GenAI deployments look like in practice? Here, we take a look at five ways that organizations can transform GenAI into an enabler, not an adversary, of environmental, societal, and governance, benefitting not only the companies themselves, but also society at large.
First, let’s be clear: Considering AI’s extensive environmental footprint, the way organizations leverage the technology will impact their overall sustainability and, with that, their attractiveness in the eyes of investors and the public. Companies can seize on this to stand out against competition, for instance, by partnering with renewable-energy-powered data centers. By incorporating AI into an overarching ESG strategy, companies can be leaders in both technological innovation and corporate sustainability.
The complexity of large language models (LLMs) is directly correlated with their environmental impact, typically dwarfing that of the online technologies we’ve relied on until now. According to a report by Goldman Sachs, a single ChatGPT query consumes nearly 10x the electricity of a standard Google search. By considering trade-offs between performance and sustainability, as well as the cost of training LLMs, companies can effectively manage the carbon footprint of their AI deployments.
Not all AI tasks require the most powerful, resource-intensive large language models. For many use cases, smaller, less complex models can deliver comparable performance with far lower energy consumption. If, for example, a smaller language model (SLM) can accomplish a given objective just as efficiently as a larger model, it can be a smart choice that also helps meet sustainability objectives. Careful scoping of AI-enabled tasks can help companies optimize deployments to avoid unnecessary energy consumption and keep their carbon footprint in check.
As AI matures, organizations need to be ready to adapt to new technological opportunities and fast-changing regulatory requirements. One way to future-proof AI deployments is through solutions that are LLM-agnostic, meaning they are not tied to any specific AI vendor or technology. This not only lets companies more easily pivot in their AI strategy as newer, more energy-efficient models emerge, or as regulations around AI and sustainability become stricter.
Finally, companies can leverage efficiency gains achieved using AI to offset their environmental footprint associated with the technology. AI can, for instance, be used to optimize supply chains, reducing waste and energy consumption across various operations. It can also help improve energy efficiency in manufacturing, identifying areas where resources are wasted and proposing real-time adjustments.
The Squirro Enterprise GenAI Platform offers a variety of features that help organizations align their AI deployments with ESG goals while also maximizing performance and efficiency of large-scale deployments. Here are some of its highlights:
Learn more about the technology behind the Squirro Enterprise GenAI Platform in our Technical Essentials Guide for Knowledge Management. And if you’d like to learn more about how the Squirro Enterprise GenAI Platform can enhance your business, book a free demo!