How AI could boost GDP and help reduce greenhouse gas emissions

The application of AI technologies in four areas – agriculture, water, energy and transport – have the potential to increase global GDP by up to $5.2 trillion by 2030, according to a new report from Microsoft and Pricewaterhouse Coopers. That is an increase of 4.4% in global GDP over the next 11 years, relative to business as usual.

At the same time, these technologies could reduce global greenhouse gas emissions by up to 4%. That is equivalent to the predicted 2030 annual emissions of Australia, Canada and Japan combined.

This map shows where those changes could occur.

Summary of regional GDP and GHG impacts relative to the baseline by 2030 in the “Expansion” Scenario

Europe could see the greatest rise in GDP – an increase of 5.4%, while the United States could see the greatest fall in greenhouse gas emissions – a drop of 6.1%.

The report also predicts that, without addressing some of the blockers to technology and AI adoption and readiness, the impact of AI will be felt less in some parts of the world. If progress is made on those blockers, however, these regions could benefit greatly from low-carbon, sustainable economic growth.

[Subscribe to Microsoft on the Issues for more on the topics that matter most.]

Of the four sectors detailed in the report, the impact of AI within the energy and transport sectors is predicted to contribute most to the rise in GDP and to the fall in harmful emissions. The following shows the changes forecast in each of the sectors.

Key sectoral results – impact on GDP and GHG emissions by 2030 in the “Expansion” scenario

The high potential of transformation within energy and transport sectors is due to an array of innovations, some of which are already being realized including traffic optimization systems, vehicle-sharing services, increased efficiency of renewables and the smart management of energy consumption, in two heavy-emitting industries.

For more on AI innovations at Microsoft follow @MSFTIssues on Twitter.