While seemingly every industry is developing artificial intelligence tools to streamline processes and improve efficiency, few industries have a better use case than carbon accounting. The flexibility of new AI tools, the ability to analyze extensive data sets, and the power to make accurate estimates when necessary all demonstrate the importance of new computing tools in this sector. By utilizing computing power to track supply chain and product emissions, companies will gain a better understanding of total emissions, where these emissions originate, and what policies can be implemented to reduce them.
The pressure to develop tools for carbon accounting varies across different regions. New carbon border adjustment mechanisms (CBAM) in Europe require importers to the continent to provide accurate carbon reporting data by 2026. Emissions accounting is in growing demand, as in 2022, only around 10% of emissions were measured by companies, and, unsurprisingly, only around 14% of companies were able to reduce emissions at a pace matching their benchmarks. Whether as a matter of principle, as a response to shareholder pressure, or to meet new regulations, companies will need to start taking carbon emissions more seriously.
In order to understand the complexity of tracking carbon emissions, it is important to understand the distinctions. Emissions are categorized into 3 scopes. Scope 1 emissions are those that are directly controlled or owned by the company in question. For the aviation industry, this may be the fuel required by the planes. In shipping, this would entail the emissions from the fleet of diesel trucks. Scope 2 emissions are generated by the energy we purchase, such as the electricity used in an office building. The real challenge lies in calculating scope 3 emissions, which encompass the emissions generated by a company’s entire value chain. For instance, the emissions required to build all parts of a computer separately must be calculated by Apple to track emissions accurately.
Meet EP Investing — capital discovery for the energy transition.
Scope 3 emissions are significant because they tend to account for the majority of a company’s emissions. Fortunately, a Boston Consulting Group (BCG) report on carbon emission calculations found that companies are improving their ability both to measure and mitigate scope 3 emissions. It is not difficult to imagine why AI would be most impactful in measuring Scope 3 emissions, given the substantial amount of data required to accurately track emissions generated in the supply chain. The same study by BCG found that companies using automated digital solutions to measure emissions are 2.5 times more likely to measure emissions comprehensively.
Companies such as SINAI Technologies have thus far utilized AI to gain a better understanding of calculating carbon emissions from energy-intensive industries. During a conversation with Maria Carolina Fujihara, Founder and President at SINAI Technologies, she outlined both the benefits and complexities of calculating carbon emissions in sectors such as steel, manufacturing, and oil and gas. These industries possess significant physical assets, which makes the calculation of emissions a complex process. However, SINAI has trained its algorithm to read uploaded datasets and organize system information. Along with reliable calculations, this creates a system that enables SINAI to provide recommendations to clients that are more likely to be approved.
One of the key requirements for tracking both carbon emissions and green offsets is their reliability. In conversation with James Allan, Senior Director at Quinbrook Infrastructure Partners, Allan explained that companies need reliable data on carbon emissions to adhere to their reduction goals. As Allan explained, working with high-profile companies to achieve climate goals requires being able to track both emissions and offsets reliably. Allan explained how Quinbrook has created products capable of solving the tracing problem, demonstrating exactly where and how green electrons are entering the grid. The goal is to establish the long-term benefits of a zero-emission approach with strong validity.
From a technical standpoint, a major benefit of commercializing new computing tools is the flexibility it offers. Darayush Mistry, Head of Product at Pulsora, discussed how flexible AI tools can enable the creation of products for a rapidly evolving industry. As Mistry outlined, the sustainability landscape, including the policies and technologies, is constantly shifting. Therefore, the AI and machine learning tools used to calculate emissions must be able to adapt and improve continually. Additionally, using actual data whenever available is key to accurate accounting. While new technological tools can provide certain estimations, Mistry emphasized that accurate data is a cornerstone in reliable emissions accounting.
It is worth noting the actual emissions that result from using AI to calculate them. The International Energy Agency reports that by 2026, the combined use of AI, data centers, and cryptocurrency could account for 4% of global energy consumption. Utilizing renewable sources to power data centers, optimizing energy usage within the center, and purchasing offsets are crucial for any company that intends to use large amounts of computing power.
Whether in conversation with Fujihara, Allan, or Mistry, the conversation turned to the importance of demonstrating value. All three explained that companies that make a commitment to decarbonization need not see it as a cost, but rather as an opportunity. By addressing carbon emissions and reductions now, industries ranging from steel to technology will be better positioned for the future.
Meet EP Investing — the platform behind the stories.
1,300+ companies · 350+ investors · 47+ grants
