Elsevier Q&A: Circularity & the Future of Sustainable AI

The world is heavily influenced by climate urgency, shifting regulations and evolving consumer expectations. Sustainability is no longer a peripheral concern, it is central to how organisations operate, innovate and stay competitive.

For R&D-led businesses in particular, the challenge lies not only in reducing environmental impact but in developing new materials, technologies and processes that enable long-term resilience and compliance.

As organisations navigate complex supply chains and growing demands for transparency, circular innovation has emerged as a critical goal.

Yet achieving it requires collaboration, shared standards and access to reliable, cross-disciplinary knowledge.

AI is becoming a vital enabler in this space, helping researchers identify greener materials, optimise manufacturing and accelerate the transition away from wasteful linear models.

Stuart Whayman, the President of Corporate Markets at Elsevier shares his expertise with Sustainability Magazine.

Please introduce yourself and your role

I’m Stuart Whayman and I’m the President of Corporate Markets at Elsevier.

I work at The Corporate Markets group at Elsevier, supporting R&D organisations with trusted scientific information, advanced AI tools and professional services.

I also serve as the Executive Sponsor of Elsevier’s Climate Action Programme and Co-Chair of its Climate Advisory Board.

Elsevier works to advance the UN Sustainable Development Goals (SDGs) through evidence-based, measurable action in our business and in partnership with the communities that we serve.

The Climate Advisory Board’s role is to advise on key actions and initiatives to deliver a net zero future and advocates for science-based decisions and action to address the urgent climate emergency.

How is the landscape of sustainability priorities evolving?

Consumers and companies both want to shift to more sustainable operating models.

However, the global economic downturn has created a challenging financial environment for businesses.

Despite one survey finding consumers would pay nearly 10% more for a sustainable service, in some cases sustainability is being deprioritised by businesses in favour of profitability.

The shift away from sustainability also poses organisations a compliance headache.

With jurisdictions introducing regulations to govern sustainability, transitioning to more sustainable operations isn’t a nice to have, it’s an obligation.

For example, regulations and guidelines around “forever chemicals,” including PFAs, are changing the ways organisations have to consider sustainability.

With these regulations in mind, it is encouraging to see that the move towards carbon-neutral production has begun as organisations look to develop greener materials and adopt more ecologically friendly manufacturing processes.

Why does circular innovation require a collaborative approach?

With how interconnected our global supply chain is, it is impossible for one individual organisation to be responsible for any product’s full lifecycle.

Partly because of this complexity, very few organisations have a solid circularity strategy in place to overcome technology or culture-related challenges.

It is a disappointing trend that we are continuing to move further away from a fully circular economy: this year’s Circularity Gap Report reported a drop in global circularity from 7.2% to 6.9%.

Different industries also have different standards for their data practices, technology and processes.

These all make implementing cross-industry circularity more difficult.

While complete standardisation across industries isn’t possible, different industries can work to closely to align their standards to promote greater circularity.

To promote cross-industry collaboration, organisations must have access to flexible information sources to expand expertise beyond their own field and understand all parts of their product lifecycle to promote greater circularity.

How can AI contribute to the development of sustainable materials?

AI is able to parse large datasets far more quickly than a human could to accelerate R&D into green materials and ecologically friendly manufacturing processes.

For example, applying AI to search large chemical databases and published literature can save significant time in materials R&D to identify shorter reaction routes, avoid hazardous materials and target synthesis routes with higher yields.

The ability to test material properties far more quickly than human researchers also allows AI tools to identify new use cases for existing materials, which is an even more efficient and sustainable use of resources than discovery of novel materials.

Many innovations in the materials sector at this stage will seem incremental but could have far-reaching consequences.

For example, masks from the COVID-19 pandemic have exposed habitats and water systems to billions of microplastics after improper disposal of vast amounts of polypropylene-based masks.

Testing large quantities of sustainable materials to identify green alternatives to plastic production is just one of the potential use cases for AI tools.

These tools are able to sift through more data than human researchers working alone, allowing them to identify materials with different properties or repurpose existing materials to reduce overall plastic production and pollution.

Could AI support accountability in sustainability?

The primary way organisations are held to account for their decisions is via regulations, which bring the threat of fines and other penalties.

AI could potentially accelerate regulatory reporting within organisations to play a role in keeping organisations compliant.

AI can be used to oversee processes and ensure full regulatory compliance with minimal potential for human error.

Of course, any outputs in crucial business areas must be verified by a human-in-the-loop, so there will always be apart for subject matter experts to play in ensuring regulatory compliance.

AI has huge energy and water demands, how can businesses use AI sustainably?

AI has a role in helping us solve some of the most urgent sustainability issues we are facing.

Nevertheless, we have to be very mindful of the significant energy requirements of AI and look at ways of making AI itself more sustainable.

Any technique that can reduce the size of a search will help to reduce AI’s energy consumption.

One such technique is leveraging more effective architecture, such as retrieval-augmented generation (RAG) architecture.

RAG frameworks reduce the search space to only relevant datasets and documents, which minimises the energy consumed per search or question posed to an LLM.

Additionally, working with data science experts to identify the most accurate and verified datasets for model training reduces the need to continue retraining models on new data.

This helps to decrease the energy cost of LLM use, shifting it closer to regular website usage.