Trump promotes AI energy agreement: OpenAI, Microsoft, Google, and others pledge to cover their own data center electricity costs

On March 5, the White House announced that several major tech companies have signed an “Energy Commitment to Protect Taxpayers,” pledging to cover the costs of electricity supply and grid upgrades for AI data centers to prevent AI infrastructure expansion from driving up residents’ electricity bills. Participating companies include Amazon, Google, Meta, Microsoft, OpenAI, Oracle, and xAI.

President Trump stated at a White House press conference that the commitment aims to ensure the rapid development of the AI industry without passing energy costs onto ordinary families. According to the agreement, these companies will provide or pay for all power generation costs for new AI data centers and bear the costs of infrastructure upgrades such as transmission lines, substations, and grid expansion.

The White House pointed out that companies must also negotiate independent electricity pricing mechanisms with utility companies and state regulators to prevent increases in residents’ and businesses’ electricity bills caused by AI infrastructure development. This move is seen as an important policy measure by the U.S. government amid surging demand for AI computing power and energy concerns.

Meanwhile, uncertainties in the global energy market are intensifying. As tensions in the Middle East persist, U.S. military actions related to Iran have raised concerns over the security of energy transportation through the Strait of Hormuz, causing fluctuations in international oil prices to become a focal point of discussion. Against the backdrop of increasing electricity demand from AI data centers, energy costs have become a key issue for the tech industry and policymakers.

Several tech companies stated during the meeting that they are supporting AI infrastructure development by adding new power generation capacity and investing in grid upgrades. xAI announced that it will support AI development by increasing electricity supply near data centers and aims to minimize the impact of energy prices on the public. Amazon emphasized that investing in data centers while simultaneously building grid infrastructure is crucial for ensuring the stability of U.S. energy supply and maintaining economic competitiveness.

Meta CEO Dina Powell said that the company plans to fully cover the energy costs of its data centers and to train technical talent related to data center construction, such as fiber optic technicians, electricians, and welders.

Google CEO Ruth Porat pointed out that the expansion of AI computing scale significantly increases energy demand. For example, in Texas, as data centers expand, Google has signed agreements for over 7,800 megawatts of new power capacity.

Policy advisor Michael Kratsios stated that the core goal of this commitment is to ensure the U.S. remains a leader in AI while preventing grid stress from translating into higher energy costs for residents. As AI computing power demand continues to grow, the electricity consumption of data centers has approached the level of small and medium-sized cities, making balancing technological development and energy costs a new industry challenge.