The opportunities AI is bringing to organisations are immense, some advertisers are looking to AI to increase productivity, AI can help consumers understand their credit scores and could contribute $115 billion to Australia’s economy if implemented correctly.
And now, AI is being seen as a potential catalyst in driving sustainability within organisations.
A BCG report highlighted that 87 percent of climate and AI leaders find AI to be a helpful tool in the fight against climate change.
From that same report, 43 percent of organisations, led by industrial goods companies, can envision using AI for climate at their organisations.
While the concept of using AI to help sustainability efforts isn’t new, Kelly Brough, managing director, applied innovation ANZ at Accenture said what has changed is the awakening and awareness within organisations of how AI can be used strategically for sustainability.
“We're starting to see it applied at scale in the corporate environment around the world, both as the topic of sustainability rises in people's not only awareness but urgency to take action,” she said.
“As the technology gets more accessible, organisations can do this in no code and low code environments in order to create positive impact.”
However, with the growing use of AI, comes the energy efficiency factor.
AI is a technology that relies on huge amounts of electricity. Bloomberg reported earlier this year that AI use at Google accounted for up to 15 percent of the tech giant's electricity bill.
Kristin Moyer at Gartner noted with generative AI alone it relies on huge models trained with massive amounts of data, making it thirsty for cooling water and hungry for electricity.
So how does that make AI sustainable?
David Hirst, group executive at Macquarie Technology Group said the AI technology itself lives in a data centre, so it depends on how the power is provided to the data centre.
“For example, in New Zealand, all power in the grid is green. When a data centre is in New Zealand, or in Canberra for that matter, the power that you are consuming is not impacting the environment from a carbon perspective,” he explained.
“But when you are in the rest of the world, the majority of the world, your data centre is plugged into power that is emitting carbon into the world.”
Due to AI needing huge amounts of power and cooling, AI does become a large emitter, Hirst noted. But it is a better alternative than powering an AI in an office or warehouse environment.
“That being said, it is extremely efficient. Putting IT equipment inside data centres compared to the alternate, which might be into an office building or a factory floor that's not efficient at all,” he said.
“You’re much better off inside data centres, but you hope the impact those AI training modules and inferences are causing will have a benefit in terms of the efficiency gains that it's being used for in those industries. So, you hope that the gains outweigh the impact.”
Greener future
AI may be generating more energy now, but Steve Quenette, CEO and founder of sustainability and AI consultancy Innate Innovation said once algorithms improve expect to see a reduction in energy.
“We're still in the baby era of AI algorithms, and algorithms when they improve, we've seen this over decades and over many of our world's innovations, they improve a thousand-fold,” he said.
He used the example of Microsoft’s ChatGPT or Meta’s LLaMA, where businesses can implement AI technology into their systems.
“You could see then there's these other things occurring that will reduce the amount of environmental impact. It will take time; it's not going to be here today. There's not going to be an explosion of carbon expended on AI in the near term,” he explained.
Enverga at Bain and Co said executives are already becoming wary of energy consumption and AI use.
“What I've seen many organisations and forward-thinking executives do nowadays is look at ways where they can deploy some of these models using greener resources,” she explained.
“For example, cloud-based servers that are basically hosted in with renewable energy sources, so sourcing data centres that are based on renewable energy.”
Enverga said she is seeing more organisations be more cognisant of how their digital strategy and AI implementation will impact the company's broader sustainability targets.
“Having a clear understanding of sustainability, the digital strategy and how you align them together, then putting people as part of that strategy is understanding that and making the trade-offs and choices,” she added.
Increase in awareness
Over the past 12 months, Estelle Parker, executive manager at the Responsible Investment Australasia (RIAA) said the relationship between AI and sustainability has witnessed a marked evolution.
“While AI has been developed for and can contribute to sustainability outcomes, and has been harnessed for climate modelling to improve predictions and combat climate change, there is also an enhanced focus on ethical considerations surrounding AI,” she said.
“Investors have grown more aware of the challenges presented by big tech, including AI. Many now believe that its consequences, both beneficial and detrimental, could surpass the creation of the internet.”
She added, “While AI offers promising advancements, particularly in sustainable investing, its integration requires careful human oversight to ensure AI-driven decisions resonate with investor values and avoid potential ethical pitfalls.”
Ethical considerations
Parker at the RIAA explained the positive and negative impacts AI has on a company’s sustainability efforts.
She said on the one hand it provides efficiency improvements, data analysis and supply chain optimisation, there are also major human rights and ethical concerns surrounding its use.
“This nexus between AI and human rights has become a key focus of the RIAA Human Rights Working Group, which has identified seven salient human rights areas for investors in the use of digital technology,” she said.
Some of these areas include privacy and data protection, freedom of expression and opinion, conflict and security, discrimination, political participation, child rights, and just transition/impact on livelihoods.
Parker added, “While AI may be able to solve complex environmental and social challenges and aid in climate change mitigation, there remain significant challenges, like data quality and biases that companies need to be aware of.”
Looking at the broader ESG picture, Parker explained the pros and cons of implementing AI to help organisations reach those targets.
“At present AI can analyse vast ESG datasets, providing insights for better decision-making and predict environmental impacts whilst ensuring transparency in governance,” she said.
“Some argue this will lead to investments which are more informed and better strategic corporate decisions.”
The accuracy of AI insights is contingent upon its data quality, which often contain significant biases,” Parker noted.
“Over-reliance on AI can diminish human oversight, potentially overlooking nuanced ESG issues,” she said.
“ESG issues to be considered and managed include cyber security and data privacy, trust, accountability, misinformation, unintended bias, human capital impacts and carbon emissions. While AI has the capacity to enhance ESG practices, careful and responsible implementation is crucial.”
Continual learning
To ensure that AI continues to benefit a business’s sustainability approach, Brough at Accenture said there needs to be constant education as the technology evolves.
“Something we talk to our clients about is you have to take a holistic mindset towards this and the cost in terms of the cash outlay as well as the cost in terms of impact on the environment and ESG commitments across the board,” she said.
“There are other impacts going beyond just the environment, in fact is one side of the equation that needs to be balanced with all of the upsides, which are pretty significant.”
When used responsibly, Brough said generative AI technologies are “quite phenomenal” and changes the game in terms of what is possible in the future.
“Like everything, we are responsible by design first, and so we have to take into account the holistic picture,” she ended.
