Key takeaways

  • By 2030, AI for the environment could boost global GDP by 4.4% and reduce global greenhouse gas emissions by 4.0%.
  • Across the water, agriculture, energy and transport industries alone, AI is forecast to generate US$5.2 trillion in the next decade.
  • Guidance and collaboration across disciplines will be needed to ensure the success of AI application to the world’s environmental problems.

The planet is facing a fight. Climate change, rapid biodiversity loss, resource depletion and food and water security are just a few of the serious issues negatively impacting the Earth’s natural systems. The need to better manage them to avert further damage intensifies daily.

A new report by PwC UK and Microsoft, How AI can enable a sustainable future, provides the first step in a larger conversation that must be had to address these problems. Examining the effect of artificial intelligence levers on GDP and greenhouse gas (GHG) emissions, the study explores the economic and environmental potential that digital transformation and decarbonisation — together — will create by 2030.

While the opportunities brought by AI will need to be critically guided to avoid threats and negotiate challenges, the benefits from its application to society’s environmental problems are vast. Focusing on four sectors that together represent around three-fifths of GHG emissions1 and which are critical to the environment, the report looks at the new approaches needed to mitigate and adapt to climate change in gradual and higher impact expansion scenarios.

The
potential

The fourth industrial revolution, brought about by emerging technologies such as AI, is reshaping industries and with them the economy as we know it. The productivity benefits of AI applications — such as automation, optimisation and higher output – across the four sectors examined could boost global GDP by 3.1-4.4%, a potential yield of US$3.6-5.2 trillion by 2030. Europe, East Asia and North America could potentially each achieve GDP gains in excess of US$1 trillion. Workers too, could reap the rewards of environmental AI application, with modelling suggesting the creation of 18.4-38.2 million net jobs globally.

These economic gains are themselves reason for positivity, but as the report makes clear, “the AI for environment opportunity at its core is about harnessing one of the most powerful tools humans have created to counter environmental degradation, contribute to a low carbon transition, and help protect our planet”.

Alongside the economic benefits, the report suggests that AI could contribute to the transition into a low-carbon world, reducing GHG emissions by 0.9-2.4 gigatonnes of CO2 equivalent. Further, AI-based early warning systems for investigating illegal deforestation could save 32 million hectares of forest in the next ten years. Accurate, localised air quality warnings have the potential to reduce the risk — and cost — of healthcare impacts in the same timeframe, a possible global economic benefit of US$150 million.

Agriculture

To feed the world’s population by 2050, current agricultural production needs to more than double.2 With the amount of land available for farming already almost at capacity, and farming practices under the spotlight for their ecological impact, the land we have must produce more and cause less damage.3

Many of the digital innovations in agriculture are underpinned by AI. From robots that can pick, plant and protect crops autonomously, precision monitoring of plants and livestock to increase yield and reduce waste (such as mass spraying of pesticide or fertiliser) all the way to better land-use management, agritech promises great changes to modern farming practices. Through a reduction of fossil fuel use and greenhouse emissions from optimisation of farms and land, AI levers could increase global GDP by up to 0.2-0.3% and reduce global emissions by 0.1-0.3% relative to today’s baseline.

AI for agriculture

Energy

Energy production remains a challenge for environmental stability. The dream of reliable, affordable energy with minimal negative impact is one that could be achieved with AI — improving efficiencies and reducing the dependence on fossil-fuel.

The application of AI to the energy sector offers the chance to optimise consumption via monitoring and maintenance, and predict short and long term fluctuations. Decentralised local energy networks, operated by automated processes, promise to improve efficiency and reduce waste. And of course, AI could be beneficial in increasing the efficiency of renewable energy assets alongside those of fossil fuels.

In the energy sector alone, these levers could deliver a 1.6-2.2% rise in global GDP by 2030. And by the same year, render a drop in GHG emissions of 1.6-2.2%, the greatest potential reduction in all four sectors.

AI for energy

Transport

Safe, efficient and sustainable transport — for cargo and people — is an important goal for environmental health as 20-30% of global energy consumption and CO2 emissions are attributed to transport.4 Autonomous and semi-autonomous vehicles, if taken up extensively as in the report’s ‘Expansion’ scenario, could greatly decrease GHG impacts. Autonomous deliveries, from long-haul trucks to last mile robots would also contribute to emissions reductions.

Other AI applications that could be implemented include optimised traffic flows, eco-friendly driving features and real-time smart pricing for tolls (such as congestion charges), better forecasting for traffic prediction and logistics planning, and again, the maintenance of vehicles and parts.

Together, this could generate a 1.2-1.8% boost in global GDP, and effect greenhouse gas emissions by +0.3 to -1.7%.* Moreover, these levers are predicted to offer substantial cost savings related to labour and capital in the sector because of their associated automation effects (for example, traffic optimisation). In turn, this automation would support the expansion of other sectors.

AI for transport

Water

Similar to the potential food shortages in the agricultural sector, the availability of freshwater is also expected to fall short. Global demand for water is expected to exceed supply by 40% by 2030.5 Pollution, urbanisation and climate change are all affecting the global water cycle, but AI could lessen the impact by reducing waste.

With AI, water infrastructure can be monitored in real time, predicting faults before they happen, and identifying maintenance to optimise systems. Monitoring tools can track actual water use by both business and private households, allowing suppliers to reduce both waste and predict shortages. Finally, water treatment and desalination processes could be optimised via AI, allowing for the reuse of greywater.

Such applications would have an (0.04-0.2%) impact on global GDP by 2030.**

AI for water

Challenges and
opportunities

Say the authors: “As with any new wave of technology, alongside huge opportunities, challenges need to be identified and addressed for AI’s full potential to be realised for the environment.”

Be it government or private sector, a set of challenges must be surmounted, from governance, resourcing and deployment, to collaboration and the maturity of data and infrastructure. Awareness and engagement will be key, and needs to be prioritised in a number of sectors. Multi-disciplinary collaboration between governments, tech, industry, NGOs, and academia will similarly be critical to overcoming issues.

The scale of the environmental challenges faced by people and the planet are vast, complex and urgent: from increasing climate change and biodiversity loss, through to pollution and acidification of our oceans. The rapid evolution of AI, however, presents a new opportunity to tackle these challenges with much more powerful and precise tools than ever before.

It’s time to use them for the sake of the planet.


* The report notes that energy demand may increase or decrease depending on a number of behavioural factors, for instance, the value of the convenience of not having to drive, or the market share of electric cars.

** The impact on GHG emissions of water levers assessed is negligible.

For further information on how AI can be used to enable a sustainable future, including regional analysis and recommendations for implementation, download the full report,  How AI can enable sustainable future.

 

Celine Herweijer

Contributor

Celine Herweijer

Celine Herweijer is a Partner in the firm and leads PwC’s Innovation and Sustainability work, in addition to  PwC’s climate change and international development business.

More About Celine Herweijer
Benjamin Combes

Contributor

Benjamin Combes

Ben Combes is an Assistant Director at PwC and a senior economist in the Sustainability and Climate Change team, currently focussed on harnessing emerging technologies for the environment.

More About Benjamin Combes
Jonathan Gillham

Contributor

Jonathan Gillham

Jonathan Gillham is a director of Econometrics and Economic Modelling for PwC United Kingdom.

More About Jonathan Gillham