Emerald AI, Google and NVIDIA have launched the AI Energy Management Alliance (AEMA), a new industry group focused on making large AI datacentres more responsive to electricity grid conditions.
The alliance brings together companies from across the AI, semiconductor, energy and utility sectors. Its 18 launch partners include Anthropic, Analog Devices, AES, National Grid, RWE, Constellation and NRG, alongside energy storage, software and infrastructure companies.
AEMA’s central focus is demand flexibility: allowing data centres to adjust their electricity consumption in response to grid conditions, particularly during periods of high demand or constrained supply. The group argues that this could allow some new facilities to connect to existing networks without relying entirely on additional generation and transmission capacity.
Potential approaches include co-located generation and battery storage, as well as shifting or reducing computing workloads when the grid is under stress. The alliance also plans to advocate for changes to interconnection and energy market rules that recognise the contribution of flexible electricity demand.
Varun Sivaram, CEO of Emerald AI, said flexible data centres could connect “faster and at far greater scale” while supporting local grid reliability.
For Google, the initiative is part of a wider effort to manage the electricity requirements associated with rapidly expanding datacentre capacity. Tyler Norris, Google’s head of advanced energy market innovation, said demand flexibility could help increase the speed at which new facilities obtain power connections.
NVIDIA said it would support technology-neutral standards based on measurable grid-reliability performance.
Frank Lacey, an energy industry executive, has been appointed AEMA’s executive director. The organisation is described as a successor to the Advanced Energy Management Alliance, established in 2014.
The launch comes as datacentre developers face growing challenges in securing electricity connections in areas where grid capacity is already constrained. AEMA’s members will now work on technical and policy approaches intended to make large, flexible loads easier to integrate into existing power systems.
