3 ways AI can drive your sustainability goals in 2024

Organisations today are expected to not just meet their sustainability requirements but also find new opportunities for positive change.

In brief

• AI analysis of new reporting directives ensure businesses are better prepared to fulfil their CSRD, CSDD and CBAM requirements.
• By collating information from multiple sources, AI can assist organisations with next actions and extract best practices from market-leading performers.
  
• Provision of detailed sentiment analysis means companies understand how their sustainability performance is perceived and what changes they need to make.
  

Last month the World Economic Forum reiterated the need for urgent action on climate change, also the core message from COP28. With the world poised at this make-or-break moment, societal and stakeholder expectations of the role companies have in reducing the effects of climate change are at an all-time high.

A U.S. Pew Research Centre Survey last October found that 52% of respondents believe large businesses and corporations can do "a lot" to reduce the effects of climate change. It indicates that the expectation has moved beyond businesses just fulfilling their ESG responsibilities to a belief that they should be focused on even greater change. The latter - termed “regeneration” – calls for a reinvention of systems across an organisation, from business models to supply chains, to drive positive impact, not simply to avoid a negative one.

But while this is certainly an important objective, many organisations currently are faced with external and internal pressures, long-term planning challenges, and reporting requirements that have grown in scope and complexity, to even reach a stage of compliance and organisation, let alone regeneration.

It’s here that Artificial Intelligence (AI) is a game-changer. By harnessing data and driving efficiency it can help your organisation meet your most immediate sustainability goals: achieving carbon neutrality, reduction of water use, and meeting SBTI targets and UN Sustainable Development goals. At the same time, AI also frees up your people to consider the bigger, long-term regeneration opportunities that can truly change your organisation’s environmental impact.

Here is 3 ways AI can assist with your sustainability goals today:

One: Guidance on sustainability reporting standards

New directives such as the Corporate Sustainability Reporting Directive (CSRD) - which is coming in this year - and Corporate Sustainability Due Diligence (CSDD) mean companies are facing increasing reporting requirements. The high volume of reporting points and the interrelationships between regulatory reports and voluntary frameworks (GRI, SASB & CDP) adds to the complexity of the task and requires organisations to be able to interpret complex policy documents in a short space of time.

Unsurprisingly, many organisations are struggling with where to begin, unsure of how they fare compared to expectations, and are confused by the multitude of requirements. As a result, they are unable to forge an action plan or identify potential problems.

Generative AI can alleviate this concern. It’s ability to analyse large volumes of documents (in this case the reporting requirements and frameworks) in real-time and then to provide easy-to-understand explanations gives companies a clear starting point. It also cuts down on complicated, manual research time and ensures consistency in understanding and actions amongst staff.

A chatbot is one means of achieving this. It can ingest all the legislation, directives, frameworks, and facts relevant to your company’s sustainability needs and then act as a “personal assistant” for any user questions. By combining “knowledge” from a vast number of resources, your organisation-specific chatbot can provide enhanced understanding on complex topics at speed, support decision-making, and even provide references so users can review the sources or answers for fact checking and traceability.

Two: Actionable insights

With the objective to halve emissions by 2030, companies must have a comprehensive and integrated net zero approach involving all aspects of their operations and value chain.

But while this integrated approach is key to meeting targets, extracting information from multiple sources and the analysis of that information (crucial if opportunities and hot spots are to be identified quickly and adjustments made) means considerable work for teams.

GenAI has the ability to monitor and analyse multiple data points, often combined with outputs from ML or other algorithms, fast and efficiently (e.g., for forecasting of total emissions or identification of raw materials that have the highest impact on CO2 reduction). It can also     enhance the quality of insights generated by this analysis by providing explainable and clear “next best actions.” For example, instead of reading the outputs of multiple reports, dashboards, and models, a GenAI powered system that is “aware” of all these data points could provide suggested scenarios of actions tailored to a stakeholder group’s objectives. This includes:

• Supply chain optimisation in which you can automatically rank your suppliers based on sustainable criteria, like their carbon footprint, water usage, and ethical labour practices.
• Driving energy efficiency opportunities by identifying underperforming assets and processes that consume excessive energy.
• Identification of processes or locations with excessive water use or risk of impacting local water scarcity.
• Benchmarking your ESG performance against your peers, enabling faster understanding and action. It also can be used to monitor for adverse events in the value chain and highlight anomalies in ethical reports thereby reducing risk and ensuring your business remains compliant.
  

Three: Sentiment analysis

Public sentiment can significantly impact a business's reputation and performance. Social media in particular - a key source of sentiment information with many people sharing their views and experiences - can often prove difficult for companies to monitor and manage quickly.

Sentiment analysis can assist with this. A form of Natural Language Processing (NLP) that uses AI to evaluate and classify sentiments expressed in textual data can provide consolidated insights to businesses. Up until recently sentiment analysis required extensive training data which made the process time consuming and expensive. Now, with the emergence of LLMs (Large Language Models), the process has been revolutionised.

LLMs perform very well when it comes to classifying text and analysing sentiment without the need for prior training, thus streamlining the sentiment analysis process. This innovation makes the collection and interpretation of public sentiment more seamless, helping businesses get a quicker and more accurate understanding of how they are perceived by the public.