Tech developers are using AI to estimate emissions, report to govt

Big banks, insurers and agribusinesses are soon expected to be required to report their farm and land emisssions. Tech companies are using AI to estimate their exposure.

Artificial intelligence is helping banks and large agricultural funds estimate their clients or portfolios’ farm emissions and climate exposure - and the tech could soon be used to report the results to governments.

It comes after legislation proposing mandatory climate-related disclosure was introduced to federal parliament earlier this year.

More than 6000 entities, including big banks, are initially expected to be required to report climate data from January next year, with other large businesses phased in after that.

Australian Securities and Investments Commission chair Joe Longo recently said the growing interest in environmental, social, and governance issues was driving “the biggest changes to financial reporting and disclosure standards in a generation” and was a “transformational issue for global markets”.

Sydney-based technology developer, WollemAI chief executive officer Sam Sneddon, said it was expected that large entities would be required by government to report Scope 3 emissions from land and agriculture portfolios in the second year of reporting.

WollemAI was the “world’s first” company to develop AI and machine learning to measure and report emissions from land and agriculture, Ms Sneddon said.

“Our customers are large organisations with agriculture land portfolios,” she said.

The software produced “accurate and actionable audit grade emissions and physical risk metrics”, she said, and could also be used to model how different farming practices withstood climate variations.

“If you are a banker, insurer, agrifood or agribusiness, you want to understand how vulnerable you are to extreme weather, or regulation, and that needs to be an accurate understanding at the farm level,” she said.

AI was used to “fill in the gaps” by “completing missing data sets” on how climate and management changes could impact different systems, she said.

In a beef grazing enterprise, the software could be used to simulate a virtual copy of the farm and model how different pasture management could affect the herd, emissions and production.

“We can run simulations for large asset owners to understand future (possible) scenarios, and stress tests,” she said.

This could be used to make emissions estimates, or assess carbon stocks.

“A lot of people are looking at Stage 3 emissions measurements but they haven’t tackled ag because it is very hard... it is a very complex space,” Ms Sneddon said.

Along with the AI employed, WollemAI now employs eight people, including agronomists and climate scientists, she said.

Sam Sneddon