Artificial Intelligence (AI) has taken the world by storm in 2023.

You may not be aware, but AI has been around since the 1950s and has been used by businesses for decades. In November 2022, the use of AI shifted to the mainstream market when OpenAI released ChatGPT, with the platform now having over 180.5 million users.

AI simulates human intelligence processes using computers and uses data centres and generators to power AI decision-making.

Undoubtedly, AI has innumerable benefits for businesses. American Express uses AI to analyse customer data and drive better lead gen processes to scale revenue. Amazon has used AI in Seattle, using cameras to scan customers and their interactions with products to eliminate the need for checkouts – simply grab and go. This drives cost efficiencies by reducing the need for store personnel.

Despite the benefits of AI, however, concerning environmental, social, and governance (ESG) issues, AI has become a topic of contentious debate. Areas such as energy consumption, workers' rights and human rights, and effective governance on the use of AI all come into the discussion.

Starting to see the bigger picture? Let’s delve a bit deeper.

Environmental concerns

AI relies on highly demanding processes from computers, frequently powered by the grid using non-renewable energy sources.

It’s estimated that a ChatGPT search uses 0.001 – 0.01 kWh of energy. In comparison, a Google Search uses 0.0003 kWh. This means AI typically consumes 1,567% more energy than a simple Google search.

Given the current 180 million users, if every user performed one search a day, that equates to 1,800,000 kWh or an estimated 419.4 tCO2e. To put things into perspective, the entire country of Greenland produced only 1,530 tCO2e, meaning every user searching on ChatGPT accounts for 27.41% of the entire footprint of Greenland.

It doesn’t stop there – servers that AI relies upon typically use water cooling, and the water consumption of AI is a considerable concern in its use.

From an environmental perspective, AI has two primary vices – its overconsumption of energy and over-generation of carbon, and its high consumption of water for cooling, which directly impacts the natural environment.

Organisations using AI then need to incorporate this footprint and impact into their Scope 3 emissions and find ways to account for and reduce that impact where possible.

Despite the gloomy prospect, significant opportunities for reducing the AI footprint lie in the grid shifting toward renewable energy production. With COP28 signalling the “beginning of the end” for fossil fuel use, it’s hopeful that a renewable grid could become a reality

Social concerns

Socially, concerns exist for AI in terms of workers' rights. Many employees fear that AI could replace their job roles, with some roles around administration of data and writing becoming redundant due to the rise of AI.

It may sound ridiculous, but this practice is already beginning, with a 2023 Capterra survey of 300 HR professionals stating they plan to use AI algorithms and software to reduce labour costs.

Ethically, businesses need to consider the impact this has on the lives of their stakeholders. The practice of replacing staff with AI seemingly reinforces the concept that social value has become too quantitatively focused and does not consider the people behind the numbers.

When it comes to AI, organisations need to ensure its ethical use, avoiding job role replacements and focusing on augmenting and improving efficiencies. The emphasis should be on training staff and developing roles, not replacing them.

Governance concerns

Regarding AI firms themselves, their governance structures are interesting. OpenAI operates as a non-profit and aims to “ensure that artificial general intelligence (AGI)…benefits all of humanity”. With a primary fiduciary duty to humanity, the duty to support humanity takes precedence over profit.

This structure is deliberate to ensure that profit maximisation does not lead AI into unethical territory. If the organisation chooses safety over profits, investors and executives can protest, but they cannot ultimately change the position or intervene against the board, ensuring the ethical proliferation of AI.

For organisations using AI, governance concerns exist regarding the overall use of the technology. How is it used in the organisation? How are people and jobs protected? What practices are allowed? What are not? What central body exists to mandate and report on its use?

All uses of AI in companies require good governance, as with any organisational concern. It’s imperative that when reporting on ESG, AI’s use is considered throughout, and concerns arising from environmental and social perspectives due to AI usage are governed by a central body to regulate and control its impact.

Conclusion

In conclusion, the intersection of Environmental, Social, and Governance (ESG) with Artificial Intelligence (AI) is complex. AI transforms industries, but its environmental impact, particularly energy use and carbon emissions, raises concerns. Urgent action is needed to address these issues.

Socially, job displacement due to AI-driven automation is a pressing issue. Organisations must balance efficiency gains with preserving the workforce, emphasising AI as a tool to augment human capabilities.

Governance, both at AI development organisations and those integrating AI, is critical for responsible ESG practices. OpenAI's non-profit structure prioritises ethical AI, setting a precedent. A robust governance framework is essential for ethical AI use, job protection, and transparent practices aligning with broader ESG goals.

Looking ahead, prospects for mitigating AI's environmental impact seem hopeful with a shift to renewable energy sources. Organisations must actively contribute to reducing the environmental and social footprint of AI, balancing technological advancements with ethical considerations for a sustainable future.