The opportunities and risks of AI have been one of the eyebrow-raising topics in business lately. Applications based on Large Language Models (LLM), especially ChatGPT, have rapidly changed the ways in which text-based information is sought, analysed, and generated. The analysis and production of visual data through computer vision are advancing, and have opened possibilities, for example, in the identification of diseases from magnetic resonance images. Here are a couple of examples of the opportunities these developments offer in promoting sustainable business.
Identification of Material Impacts
To enhance a business’ sustainability, the first step is to understand and describe its material impacts to nature, climate, and societies. Following this description, a comprehensive evaluation is essential to direct actions effectively, a process known as materiality assessment in reporting terms. Within this realm, supply chain management poses the greatest challenge. Understanding the impacts across diverse stages of an extensive value chain and prioritizing these for actionable steps and meaningful reporting becomes a complex task.
Stakeholder engagement traditionally serves as a cornerstone in this process, playing a crucial role in fostering transparency, trust, and brand. However, challenges arise in leveraging stakeholders’ perspectives to identify the most pivotal impacts within a company’s value chain, often due to their limited understanding of the entire value chain. Furthermore, prioritizing subjective views from stakeholders is difficult. It is also clear, that materiality assessments should be grounded in science, but for individual companies, analysing a ton of dispersed research data is challenging.
Here, AI could prove very helpful, as illustrated by the practical example of the Finnish company Upright. Upright utilizes language models to understand various impacts of value chains. The models use an extensive database of research articles, from which the model seeks and combines relevant information using phrases that describe the company’s business and products as input. For example, in conducting double materiality assessments required for CSRD reporting (i.e. analysis of a company’s impacts on society and the environment, and vice versa), such an approach, combined with stakeholder engagement, could be effective in the case of complex value chains.
Understanding Impacts on Nature through Imagery
In mapping and monitoring of natural resources, satellite images and machine learning have long been utilized, but more developed AI models and cloud technologies have opened up opportunities to the further utilisation of continuously updated and more accurate data sources e.g. in monitoring the health of forests. The quantity and quality of training data are growing rapidly, and in the future, self-learning computer vision models will likely be able to indicate what should be done in forests and other ecosystems to ensure sustainability of management of natural resources. However, cognitive abilities are still needed in gathering the best data sets, training models, formulating the right research questions and prompts, and verifying results. Additionally, the social aspects of sustainability, which are invariably linked to the use of natural resources, and are difficult to assess and promote without genuine dialogue among people.
Copilots in BI Tools
Business Intelligence (BI) tools serve as the backbone for companies and business processes in collecting, analysing, and visualizing data. They are typically the foundation for monitoring and reporting, allowing for benchmarking of their own metrics against other processes, competitors, or industries. Advanced Copilots and other AI-based tools that learn to find dependencies in the input data and recommend measures supporting both sustainability and business development are likely to evolve as part of these tools.
Opportunities Exist, but…
We need major changes in business models and our consumption habits that drive them, to ensure the vitality of our planet for future generations. Let’s hope that artificial intelligence can help us overcome our natural stupidity, and enhance our own intelligence in how we treat this planet and its resources. The slowness of actions is often attributed to a lack of data or understanding. In reality, there is already an immense amount of information, and by combining our own cognitive abilities and artificial intelligence appropriately, actions can be taken now. There is no time or need to wait for better data or models; the most important thing is to recognize and acknowledge the need for change and act now.