Kevin O'Leary is planning to build a massive AI data centre in rural Utah despite significant community backlash [1, 2].
The project highlights the growing tension between the rapid expansion of artificial intelligence infrastructure and the capacity of local environments and power grids to sustain them.
Local officials approved the project earlier this month, sparking protests from hundreds of residents [4]. Opponents argue that the facility could strain the local electricity grid, and increase energy costs for the surrounding community [5].
O'Leary said those opposing the project are "professional protesters" [3]. He said the Alliance for a Better Utah is Chinese-funded and intends to slow the development of data centres in the U.S. [3].
To address concerns regarding the power grid, O'Leary said the project includes a dedicated energy source. "We’re building a plant to power the data centre so it won’t strain the local grid or hike electricity costs for nearby residents," he said [5].
While the exact cost of the Utah facility varies by report, one estimate places the project at $100 billion [3]. This follows O'Leary's other ambitions in the sector, including a proposed AI data-centre campus in northern Alberta, Canada [1]. That Canadian project is estimated to cost $70 billion [1] and would feature a power capacity of 7.5 gigawatts [1].
O'Leary said the Utah centre is necessary to meet the increasing global demand for AI compute power [5].
“"They’re ‘professional protesters.’"”
This conflict underscores a broader global trend where the physical requirements of AI—specifically immense land and electricity—clash with local zoning and environmental protections. As investors like O'Leary push for gigawatt-scale campuses, the viability of these projects will likely depend on whether developers can provide independent energy solutions to avoid destabilizing municipal grids.
