The disruptive and transformative role of AI in the circular economy transition

This article explores key disruptive impacts of AI on the circular economy transition, followed by ways AI can positively transform circularity to provide guidance to professionals in sustainability, supply chain management, and industrial sectors to better understand the role of AI in shaping a sustainable and circular future.

Generative AI is a model which focusses on creating content such as text or images. Currently very much on trend, the breadth of technologies which fall under the term are often forgotten. Broadly speaking, AI refers to a set of technologies which are designed to allow machines to perform historically human tasks such as reasoning, learning, and pattern recognition, based on vast swaths of data such as videos, images and text.

Within a circular economy context, AI has the potential to optimise resource efficiency, improve supply chain transparency, and create new business models. Such capabilities can be delivered using a mix of AI applications. For example, a supply chain management tool might integrate computer vision to monitor warehouse inventory, natural language processing (NLP) to parse and analyse supplier contracts, social media sentiment, and user reviews, etc., and reinforcement learning to optimise logistics routes.

However, the use of AI in a circular economy context also carries risks such as reinforcing linear economic structures, contributing to resource-intensive digital infrastructure, and increasing the risk of greenwashing.

Circular disruption - How AI may impede the circular economy transition

AI’s rapid development and adoption comes with unintended consequences that could hinder the shift to a more sustainable economic model. One major concern is the resource intensification that AI requires.

Unintended resource intensification

Large-scale machine learning models and deep learning systems demand vast computational resources, leading to increased electricity consumption and reliance on rare earth metals such as lithium and silicon. The growing energy demand of AI may offset any efficiency gains in circularity, while the increasing need for advanced computing hardware could exacerbate the issue of electronic waste.

Beyond energy consumption, the carbon footprint of training AI models remains a challenge. Companies are working on greener AI solutions, such as using renewable energy-powered data centres, water-based cooling, and the reuse and integration of waste heat into district heating systems. However, the progress is slow and AI’s sustainability remains a complex issue.

Bias toward linear economic models

Another challenge is AI’s bias toward linear economic models. Many current AI-driven optimisation tools are designed for traditional linear economies that prioritise cost reduction and efficiency rather than circularity. Supply chain AI solutions tend to emphasise just-in-time manufacturing and maximising throughput – without considering circular strategies. Inventory management and production planning, for example, often favour short-term efficiency over long-term material circularity, while AI-driven procurement systems may continue to prioritise virgin materials over recycled alternatives due to cost and availability.

Furthermore, AI’s reliance on historical data means that it often reinforces existing economic patterns rather than introducing transformative solutions. If AI systems are trained on datasets reflecting decades of linear production and consumption, they may fail to recognise the benefits of circular business models. Overcoming this challenge requires deliberate programming and the integration of circularity metrics into AI training and decision-making processes.

Greenwashing and misinformation risks

AI also presents a risk in the realm of greenwashing and misinformation. The ability of AI to generate sustainability reports, marketing content, and ESG disclosures creates opportunities for misleading environmental claims. AI-powered tools can produce detailed reports and sustainability assessments that appear credible but lack verifiable impact measurement. This can lead companies to create overly optimistic sustainability narratives that do not reflect real progress. Additionally, AI-driven reporting tools might be used to comply with regulations without actually addressing sustainability challenges, misleading both investors and consumers.

One key issue is that AI models are only as good as the data they are trained on. If sustainability reporting relies on biased, incomplete, or outdated data, AI-generated insights may be flawed. Ensuring transparency and accountability in AI-generated sustainability metrics is crucial for building trust and driving real change.

Over-reliance on digital solutions

An over-reliance on digital solutions is another potential disruption. Tools such as AI-powered digital twins, predictive analytics, and platforms combining AI with other technologies, provide valuable insights, but they should not replace fundamental systemic changes needed for circularity. Companies may focus on AI-driven efficiency improvements rather than restructuring business models for circularity. Local circular economies, such as community-based sharing models and repair networks, could struggle against AI-optimised globalised supply chains. Furthermore, the complexity of AI could make circular solutions less accessible for smaller businesses or developing economies, limiting broader adoption.

Circular acceleration - How AI can accelerate the circular economy transition

Despite the challenges discussed above, AI holds immense potential to drive circular economy innovations. One of the most promising areas is in optimising resource efficiency and waste reduction.

Optimised resource efficiency and waste reduction

AI can enhance predictive maintenance, ensuring industrial equipment operates efficiently and reducing premature disposal. Advanced computer vision systems can improve waste sorting, making recycling processes more effective. AI-powered supply chain analytics, which combine learning models with other AI applications such as natural language processing, can help businesses anticipate demand more accurately, reducing overproduction and waste.

Enhanced product and material circularity

AI also plays a crucial role in enhancing product and material circularity. It assists in designing products with better material selection, modularity and lifecycle tracking. Generative design algorithms help engineers develop products that are easier to repair, recycle and reuse. AI-powered material passports, or digital twins, track materials throughout a product’s lifecycle, enabling better recovery and reuse. Additionally, AI can optimise reverse logistics, making it easier for companies to collect and refurbish used products efficiently. For instance, machine learning models can be used to assess historical return data to forecast future volumes and enable improved inventory management. Adding to this, computer vision systems can be used to assess the condition of returned products to speed up collection processes and minimise human errors.

AI-driven sustainable business models

Another area where AI is driving change is through the development of sustainable business models. AI enables the expansion of Product-as-a-Service (PaaS) models, where products are leased instead of sold, ensuring they remain in circulation longer. It can do so by supporting these models with predictive analytics on appropriate maintenance cycles to extend product lifetimes. Furthermore, models can be applied to better predict user needs and product utilisation and improve the way products are recirculated within the PaaS network. Another application for such models is supporting reuse-based commerce by improving resale and refurbishment platforms. Finally, demand prediction models powered by AI can help manufacturers align production with actual consumer needs, reducing surplus production and waste.

Improved transparency and circular supply chains

AI is also improving transparency and accountability in circular supply chains. AI-powered blockchain solutions can help ensure materials and components are traceable from production to reuse. Satellite imaging and AI-based monitoring systems can detect unsustainable practices such as illegal waste disposal or deforestation, allowing for real-time intervention. AI-driven automated sustainability reporting supports compliance with regulations like the CSRD and its associated ESRS E5 Resource Use & Circular Economy, making it easier for companies to demonstrate their commitment to circularity.

Consumer engagement and behavioural change

Consumer engagement is another area where AI is making a difference. AI-driven recommendation engines encourage consumers to choose sustainable alternatives by suggesting second-hand or refurbished products. Gamified sustainability apps incentivise circular behaviours by rewarding users for responsible consumption choices. AI-powered personal assistants can also help individuals make informed purchasing decisions by offering sustainability insights based on their preferences and past behaviours.

A double-edged sword?

AI’s impact on the circular economy transition is both disruptive and transformative. While it can reinforce linear models, increase resource consumption and enable greenwashing, it also offers significant opportunities to optimise resource efficiency, enhance transparency, and create new circular business models.

Businesses and policymakers must align AI development with circular economy principles to ensure that technological advancements contribute to a sustainable future.

For businesses, the decision to adopt AI technologies to enable the transition to a circular economy, requires a balancing exercise between the benefits gained using these tools, such as improved material selection and lower material losses through supply chain management, etc., against the impacts driven by this use of technology, for instance, purchase of cloud computing or data storage and the associated emissions. Here, it also becomes critical for companies to assess whether full scale AI models are needed for the benefits being pursued or if simpler machine learning and predictive models can solve the challenges faced.

From a policymaker’s perspective, the impacts of increased energy use, the risks of workforce displacement, and other elements tied to a just transition need to be factored into the approach to policy.

Moving forward

Companies need to embrace AI strategically to accelerate their transition to a circular economy. Ignoring AI will not only place companies at a disadvantage commercially but also impede a company’s ambition with respect to the transition to a circular economy.

The complexities and rapid developments in the circular economy now require data-driven insights and robust analytics that only advanced AI systems can provide. By integrating AI with circular economy principles, companies can move beyond incremental improvements toward transformative solutions that reshape entire value chains. The potential benefits, as outlined above, are clear, but the path less so.

To navigate the path forward, companies should begin by clearly defining their circular economy objectives and identifying where AI can deliver the greatest impact.

It is important to invest in developing comprehensive data strategies that ensure AI models are trained on accurate, relevant, and unbiased datasets, reflecting circular principles rather than linear patterns. Establishing cross-functional teams that combine expertise in sustainability, technology, and operations will foster innovation and effective AI integration. Furthermore, companies should prioritise transparency and accountability, actively countering risks of greenwashing by openly communicating AI-driven sustainability outcomes and engaging stakeholders through verified impact assessments.

By taking these deliberate steps, companies will be better positioned not just to survive but thrive in the rapidly evolving circular economy landscape.