Artificial intelligence (AI) was the buzzy technology of 2024, and it looks set to define 2025 as well. The technology promises to unlock new efficiencies and increase visibility for organisations. In particular, many organisations believe that AI could be a powerful tool for taking control of their environmental impact and making critical strides towards their sustainability ambitions.
“Throughout 2025, the supply chain sector will only continue to innovate further through the use of advanced artificial intelligence,” HaulageHub Co-Founder Scott Robertson told SupplyChain Strategy. According to Robertson, the climate crisis means that “visibility and sustainability within the supply chain have never been more important.” Therefore, it’s imperative that supply chain organisations embrace AI, which he believes will “be pivotal in improving the industry’s carbon footprint.”
Robertson isn’t alone in his support for AI as an effective weapon in the fight against climate change (or in the pursuit of visibility, efficiency, and profit). Across the supply chain sector, organisations are racing to inject AI into their operations — from more traditional analytical tools to a new wave of more independent “agentic AI” that can function with even greater autonomy.
Investment in AI (generative or otherwise) throughout global supply chains is set to grow from around $600 million in 2020 to well over $51 billion by 2030.
Driving sustainability in the supply chain with AI
In many ways, the argument that AI is poised to have a positive impact on supply chains is a persuasive one. For example, let’s look at short-haul, last mile delivery networks — one of the most fraught areas of the supply chain when it comes to emissions.
Robertson notes that, as of now, empty runs — in which an unladen vehicle travels to or from a delivery — account for over 30% of all HGV miles in the UK. This contributes over 5 million tonnes of needless CO2 emissions annually. Through leveraging AI systems, he believes, supply chain organisations could reduce this figure significantly.
“The technology enables hauliers to track and measure emissions in real time, thereby allowing them to make informed and data-driven decisions to improve their sustainability efforts across the supply chain, while also reducing costs,” he explains.
With the transport industry poised for even further transformation this year, Robertson argues that “AI and machine learning are facilitating even more innovations such as autonomous trucks, advanced telematics, and the integration of electric and hydrogen-powered vehicles.”
He continues: “The supply chain sector will need to embrace these developments with open arms to ensure companies are at the forefront of a digital, more efficient, and more sustainable future. The business case for sustainable logistics is clear and the sector will only benefit from implementing AI in their processes to achieve this in 2025.”
AI for sustainability? There’s a catch, obviously
Because of course there is. Specifically, there’s a problem with the idea that AI (especially generative AI) could be a cure-all for supply chain sustainability woes. First of all, even if AI can deliver critical efficiencies and visibility that Robertson suggests, efficiency and visibility aren’t the whole battle.
Logistics organisations need to examine alternative fuels, more circular economic practices, and a collaborative and holistic approach to tackling the climate crisis. AI is only able to fight half of that battle. Not only that, but AI may also be doing more harm than good.
Since the launch of Chat-GPT and other GenAI tools, demand for data centres has skyrocketed. An industry that fought for over a decade to reign in its electricity and water usage is abandoning its climate commitments as the demands of an AI age become apparent.
“What is different about generative AI is the power density it requires,” explains Noman Bashir, lead author of an MIT impact paper released earlier this month. “Fundamentally, it is just computing, but a generative AI training cluster might consume seven or eight times more energy than a typical computing workload.”
As a result, generative AI could consume as much energy as Japan by next year. Chat-GPT alone uses power equivalent to around 180,000 US households every day, and a single conversation with Chat-GPT uses about one regular plastic bottle of drinking water. According to OpenAI, the platform (which is just one of several AI models on the market) processes about a billion queries per day.
How can a technology that is actively contributing to the climate crisis be an instrumental part in solving it? Can the efficiencies AI creates offset the damage it does? Or will AI emissions be the next big sin shuffled under the rug of “untraceable” scope 3 emissions?
- AI in Supply Chain
- Sustainability
