Net zero goals harder to meet without AI, say UK business leaders

• Two-thirds (66 per cent) of energy and technology business leaders in the UK believe AI will have a positive impact on climate over the next three years,
• But five in six major energy consumers, which include data centres, hyperscalers and tech companies, expect their energy consumption to increase in the next three years, with half of them looking for a pragmatic solution on decarbonisation.
• Grid infrastructure limitations is the top barrier for making the most of AI to support the energy transition.

The majority (86 per cent) of energy and technology business leaders in the UK believe that without AI it would take longer for their organisation to achieve its net zero goals, according to KPMG’s latest report on AI and the energy transition. Additionally, two-thirds (66 per cent) believe AI will overall have a positive impact on climate over the next three years, with a further 24 per cent believing the impact will be net neutral.

More than 1,200 senior executives, including 157 in the UK, from major companies spanning a variety of industries across 20 countries were surveyed for KPMG’s report, AI’s dual promise: Enabling positive climate outcomes and powering the energy transition[i]. The research was conducted to better understand how AI is currently being used to drive energy transition with respondents categorised as either energy producers or consumers.

When it comes to AI’s impact on the transition to renewable energy specifically, almost all (96 per cent) of respondents in the UK believe AI’s impact will be a positive force in accelerating the transition to renewable energy sources over the next three years. UK respondents believe AI could have the most transformative impact over the next three years in the following areas:

• carbon accounting with real-time emissions tracking (39 per cent)
• smart grid optimisation and outage prediction (34 per cent)
• weather pattern analysis for energy planning and real-time demand response and load forecasting (both 31 per cent).

Wafa Jafri, Lead of Energy and Natural Resources Strategy and Partner KPMG in the UK, said: “There is much discussion about AI’s negative impact on climate goals, and it is true that the technology will need a lot of energy, with overall energy demand rising. However, it is too often overlooked just how important AI will be to the transition. Overwhelmingly the businesses investing in it do not see AI as a barrier to their sustainability goals at all, but as an enabler for them. Energy efficiency has been one of the hardest nuts to crack in the energy transition so far yet is critical if we want to reach net zero goals, AI has the ability to change this on a substantial scale.” 

Increase in energy demand

Despite the perceived positive impacts that AI could have, it is increasingly expected to also have a large impact on energy demand. Almost five in six (83 per cent) energy consumers surveyed expect their energy consumption to increase in the next three years, with over half (53 per cent) expecting this to be a significant or substantial (over 25 per cent) increase and a further 12 per cent believing it will be an increase of over 100 per cent.

Currently only 8 per cent of these energy consumers believe their AI use makes up over half of their total energy usage, however, when forecasting ahead three years this number jumps to 44 per cent who expect AI to make up over half of their energy needs.

Meanwhile, 96 per cent of energy producers are expecting energy demand to increase in the next three years, with 37 per cent expecting a substantial increase (50-100 per cent) and 16 per cent expecting an exponential increase (more than 100 per cent).

Around half of energy consumers (47 per cent) expect over half of their energy needs to be met by renewables in three years’ time, despite the expected increase in demand. Over a third (35 per cent) say there is a strong preference for clean energy to meet their growth ambitions, with a further 14 per cent saying that clean energy is non-negotiable. However, almost half (47 per cent) of those respondents say they will take a balanced approach that will depend on specific circumstances.

Leanne Allen, Head of AI at KPMG in the UK, said: “It is no secret that AI consumes a lot of energy, particularly throughout model development and deployment. Both organisations and individuals need to be making a conscious effort to ensure they are being as efficient as possible in how they are operating with and training AI, as otherwise it will become a real problem for energy demand. However, when it is done responsibly any impact can be significantly reduced and AI can improve energy consumption and production. This should ultimately drive us towards a cleaner future – whether that’s through things like grid optimisation, efficiencies in the AI models themselves or many others.”

AI seen as an enabler for efficiency for energy consumers

Improving energy efficiency and driving efficiency improvements in AI models were the top two priorities for energy consumers with 58 per cent and 52 per cent respectively. Investing in new energy technologies (35 per cent) and securing clean energy supplies (26 per cent) were the other main priorities.

However, grid infrastructure limitations were seen as the biggest barrier to achieving any gains from AI, with 50 per cent of these business leaders saying this was the case. High costs (39 per cent), technical integration challenges (32 per cent), insufficient renewable energy capacity (29 per cent) and regulatory and policy barriers (27 per cent) were some of the other barriers cited.