The rollout of AI-managed electric vehicle charging could significantly cut energy bills for drivers and bring major benefits for the grid, new trials have revealed.

The research, carried out by energy research and artificial intelligence institute Centre for Net Zero (CNZ), evaluated the impact of the world’s largest AI-managed EV charging tariff to understand the role of AI in automating energy demand.

With support from the King Climate Action Initiative at the Abdul Latif Jameel Poverty Action Lab (J-PAL), researchers conducted a randomised controlled trial of over 13,000 UK households across 12 months.

The trial assessed the tariff, which dynamically charged consumers’ vehicles when wholesale electricity prices were cheap, typically corresponding to periods of higher renewable supply and lower demand.

Key findings include that AI-managed EV charging can bring down electricity bills by up to around £650 for drivers by aligning charging to times when electricity is cheap and abundant.

The managed charging tariff led to average annual bill savings of around £340 in the trial, or approximately £650 if consumers were to switch from a standard flat tariff. This can make an EV up to seven times cheaper to run than a petrol car.

The study also showed that peak household electricity use (from 16.30 to 20.30) dropped by 42%, with the entirety of EV demand shifted to off-peak hours.

Total household electricity consumption remained unchanged, instead taking place when the grid is less strained and energy prices are often cheaper, with a 50% increase in demand from 23.30 to 05.30.

The research found automation was highly accepted by drivers, with 97.7% of total electricity consumption optimised by the AI-managed tariff, while more than half of households never used manual overrides of AI controls.

Targeted consumer engagement can increase take-up of EV charging tariffs – a simple email increased tariff adoption by 3.4% amongst ‘harder-to-reach’ users: consumers not previously on tariffs and likely more representative of future EV adopters.

In total, 85% of customers who switched to the tariff remained on it for the 12-month trial period. Users are unlikely to switch away from well-designed default options that deliver convenience and savings.

EVs accounted for 20% of global new car sales in 2024; a figure set to rise sharply in the coming years. While the decarbonisation of transport is a critical part of net zero efforts, it will also contribute significantly to rising electricity consumption – with estimates that an electrified fleet in Britain could account for 15 to 20% of total demand by 2050. This surge in adoption will place new pressures on the UK’s grid infrastructure, putting it under strain during periods of peak demand. Left unmanaged, this could increase system costs, ultimately borne by consumers, yet so far, only a quarter of EV owners in Britain are estimated to engage in smart charging to help smooth out demand.

Lucy Yu, CEO of Centre for Net Zero, said the trial confirms that AI has the potential to make our energy systems work better for everyone.

“Through automation and smart technology, consumers can optimise their electricity use, saving on bills without sacrificing on comfort. Rather than adding pressure to the grid at peak times – which is often when we are most reliant on higher-cost fossil fuels – households can make the most of cheaper, greener power. The implications stretch beyond electric vehicles: consumers can be rewarded for aligning energy-intensive tasks with high renewable production, integrating clean power more effectively and saving on system costs. “

Professor Robert Metcalfe, chief economist at Centre for Net Zero and Professor at Columbia University School of International and Public Affairs, said: “The electrification of transport is central to efforts to bring down the UK’s carbon emissions. But we need to effectively manage that additional demand from EV charging, or risk putting further strain on the grid, which consumers would bear the brunt of. This research shows that targeted incentives and automation can reduce the need for costly power generation at peak times. At scale, this could help bring down bills, make better use of renewables to meet surges in demand, and reduce the need for costly system upgrades.”

Informed by the trial’s findings, CNZ has made specific recommendations to policymakers and industry leaders.

These include a call to reform electricity market signals and the firm’s experts say wholesale electricity prices should reflect the time- and location-specific cost of supply, rewarding producers and consumers for shifting demand according to the needs of the grid.

Other calls include lower non-energy charges on bills, the introduction of more dynamic network charges, targeted engagement for tariff-switching and work to deliver market-wide half-hourly settlement.

CNZ’s full white paper is online here.